DE69125660T2 - METHOD AND DEVICE FOR DETECTING AND COUNTING DIFFERENT CURRENCIES - Google Patents

Description

BACKGROUND OF THE INVENTION Scope of the invention

The present invention relates generally to currency identification. More particularly, the invention relates to a method and apparatus for automatically distinguishing and counting banknotes of different denominations using light reflection characteristics of indicia printed on the banknotes.

Description of the related state of the art

A variety of methods and devices have been used to meet the needs of automated money handling systems. At the lower end of the sophistication of the technology are systems that are only able to handle a certain type of money, such as a specific dollar denomination, while rejecting all other types of bills. At the higher end are complex systems that are able to identify, distinguish between, and automatically count multiple denominations of money.

Banknote discrimination systems typically use either magnetic sensing or optical sensing to distinguish between different banknote denominations. Magnetic sensing is based on detecting the presence or absence of magnetic ink in areas of the printed indicia on the banknote using magnetic sensors, usually ferrite core based sensors, and using the detected magnetic signals after subjecting them to analogue or digital processing as a basis for banknote discrimination. The more commonly used optical sensing method, on the other hand, is based on detecting and analyzing changes in light reflection characteristics or transmittance characteristics that occur when a currency bill is illuminated and scanned by a strip of focused light. Subsequent banknote discrimination is based on comparing sensed optical characteristics with pre-stored parameters for different banknote denominations, while calculating sufficient tolerances reflecting the differences between individual banknotes of a given denomination.

A major obstacle in implementing automated banknote discrimination systems is achieving an optimal compromise between the criteria used to adequately define the characteristic pattern for a particular currency denomination, the time required to analyze the test data and compare it to predetermined parameters to identify the banknote under test, and the speed at which successive banknotes can be mechanically passed through and scanned. Even with the use of microprocessors to process the test data resulting from scanning a banknote, a limited time is required to acquire the samples and for the process of comparing the test data to the stored parameters to identify the denomination of the banknote. Most of the optical scanning systems available today use complex algorithms to obtain a large number of reflection data samples when a banknote is scanned by an optical scanning head and to subsequently compare the data with appropriate stored parameters to identify the banknote denomination. Conventional systems require a relatively large number of optical samples per banknote scan to sufficiently discriminate between currency denominations, mainly for those denominations where the reflection patterns are not clearly distinguishable. The use of the large number of data samples slows the speed at which incoming banknotes can be scanned and, more importantly, requires a correspondingly longer period of time to process the data in accordance with the discrimination algorithm. The processing time is further increased if the system is also adapted to count identified currency denominations. The speed at which banknotes can be classified and counted is limited because real-time processing requires that the analysis of the scanned data for a banknote is completed and the banknote has been identified as belonging to a particular currency denomination and counted before the subsequent banknote is positioned on the other side and scanned by the scanhead.

A major problem associated with conventional systems is that in order to achieve the required large amount of reflective scanning data required for accurate currency discrimination, such systems are limited to scanning banknotes along the longer dimension of banknotes. Longitudinal scanning, in turn, has several inherent disadvantages including the requirement for an extended transport path to pass the banknote longitudinally over the scanning head and the additional mechanical complexity involved in accommodating the extended web and the associated means for ensuring a uniform, non-overlapping count of banknotes with the sensing surface of the scanhead.

The end result is that systems capable of accurate banknote discrimination are expensive, mechanically bulky and generally unable to perform both banknote discrimination and counting at high speeds with a high degree of accuracy.

US-A-4 179 685, which corresponds to the preamble of claim 1, discloses a money counting and evaluating device comprising a feed mechanism, a banknote transport mechanism, a stacking station and an optical scanning head arranged between a feed mechanism and a stacking station. The scanning head includes a light source and a detector and scans an area of each banknote. A scanning device scans bit information as the banknote is moved over the scanning head. A memory stores bit information of preselected areas of the banknotes of different denominations. A processing device compares the stored bit information with the scanned bit information and determines the denomination of each scanned banknote.

SUMMARY OF THE INVENTION

It is a primary object of the present invention to provide an improved apparatus for identifying and counting To create banknotes comprising a variety of monetary denominations.

It is another object of this invention to provide an improved apparatus of the above kind which is capable of effectively discriminating between banknotes of different monetary denominations and counting them at a high speed and with a high degree of accuracy.

A related object of the present invention is to provide such an improved money discriminating and counting device which is compact, economical and has a simple structure and operation.

This object is achieved by the features as set out in patent claim 1.

Further embodiments are claimed in the subclaims.

Short description of the drawings

Other objects and advantages of the invention will become apparent upon reading the following detailed description in conjunction with the drawings in which:

Fig. 1 is a functional block diagram illustrating the conceptual basis for the optical sampling and correlation method and apparatus according to the system of this invention;

Fig. 2 is a block diagram illustrating a preferred circuit arrangement for processing and correlating reflection data in accordance with the optical sampling and counting method of this invention;

Figures 3 to 8 are flow charts illustrating the sequence of operations involved in carrying out the optical sampling and correlation method;

Figs. 9A to C are graphical illustrations of representative characteristic patterns produced by optical scanning in the small dimension of a banknote;

Figures 10A-E are graphical representations of the effect produced on the correlation pattern by using the progressive process technique according to an embodiment of this invention;

Fig. 11 is a perspective view showing a money discriminating and counting device particularly adapted to and embodying the optical scanning and correlation method of this invention;

Fig. 12 is a partial perspective view illustrating the mechanism used for separating the banknotes and feeding them in a sequential manner into the transport path;

Fig. 13 is a side view of the apparatus of Fig. 11, showing the separating mechanism and the transport path;

Fig. 14 is a side view of the device of Fig. 11 showing details of the drive mechanism;

Fig. 15 is a plan view of the money discriminating and counting device shown in Figs. 11 to 14;

Fig. 16 is a side view in section showing the angular arrangement of the photodiodes in the scanning head;

Fig. 17 is a diagram of the light distribution generated around the optical pickup head; and

Fig. 18 is a top view of the optical mask used to accommodate scanning stripes of different dimensions.

While the invention is susceptible to various modifications and alternative embodiments, specific embodiments are shown by way of example in the drawings and described in detail. It should be understood, however, that it is not intended to limit the invention to the particular embodiments disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the invention as defined by the appended claims.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to Figure 1, there is shown a functional block diagram illustrating the optical scanning and correlation system according to this invention. The system includes a bill receiving station 12 where stacks of bills to be identified and counted are positioned. The received bills are subjected to the action of a bill separating station 14 which operates to select or separate a banknote for sequential passage through a banknote transport mechanism 16 according to a precisely predetermined transport path over an optical scanning head 18 where the currency denomination of the banknote is scanned, identified and counted. The scanned banknote is then transported to a banknote stacking station 20 where the banknotes thus processed are stacked for subsequent withdrawal.

The optical scanning head 18 includes at least one light source 22 which directs a beam of coherent light downwardly onto the banknote transport path to illuminate a substantially rectangular strip of light 24 on a banknote 17 positioned on a transport path below the scanning head 18. The light reflected off the illuminated strip 24 is sensed by a photodetector 26 positioned directly above the strip. The analog output of the photodetector 26 is converted to a digital signal by an analog-to-digital converter (ADC) unit 28, the output of which is fed to a digital input of a central processing unit (CPU) 30.

According to a feature of this invention, the bill transport path is defined in such a manner that the transport mechanism 16 moves the bills with the narrow dimension "W" of the bills arranged parallel to the transport path and the scanning direction. Thus, as a bill 17 moves along the transport path on the scanning head 18, the coherent light beam 24 effectively scans the bill across the narrow dimension "W" of the bill. Preferably, the transport path is arranged so that the bill 17 is scanned approximately around the central portion of the bill along its narrow dimension, as best shown in Fig. 1.

The scanning head 18 functions to detect light reflected from the banknote as it moves across the illuminated strip of light 24 and to provide an analog representation of the change in the light so reflected, which in turn represents the change in the "black" and "white" content of the printed patterns or indicia on the surface of the banknote. This change in light reflected from the scanning narrow dimension of the banknote serves as a measure for distinguishing with a high degree of certainty between a plurality of currency denominations for which the system of this invention is programmed to operate.

A series of detected reflectance signals are obtained across the narrow dimension of the banknote or across a selected segment thereof, and the resulting analog signals are digitized by the control of CPU 30 to provide a fixed number of digital reflectance data samples. The data samples are then subjected to a digitization process which includes a normalization routine for processing the sampled data for improved correlation and for smoothing out the variations due to "contrast" variations in the printed pattern existing on the banknote surface. The normalized reflectance data thus digitized represents a characteristic pattern that is precisely unique to a given banknote denomination and provides sufficient distinguishing features between characteristic patterns for different currency denominations, as will be explained in more detail below.

To ensure complete correspondence between reflection samples obtained by scanning the narrow dimension of successive banknotes, the start of the reflection scanning process is preferably controlled by the CPU 30 by means of an optical encoder 32 which is connected to the banknote transport mechanism 16 and precisely follows the physical movement of the banknote 17 over the scanning head 18. More specifically, the optical encoder 32 is coupled to the rotational movement of the drive motor which generates the movement imparted to the banknote as it is advanced along the transport path. In addition, it is ensured that positive contact is maintained between the banknote and the transport path, particularly when the banknote is being scanned by the scanning head 18. Under these conditions, the optical encoder is able to precisely follow the movement of the banknote relative to the light strip generated by the scanning head by monitoring the rotational movement of the drive motor.

The output of the photodetector 26 is monitored by the CPU 30 to initially detect the presence of the banknote under the scanhead and subsequently to detect the starting point of the printed pattern on the banknote, which is represented by the thin borderline 17A which typically encloses the printed identification feature on the banknotes. Once the borderline 17A has been detected, the optical encoder is used to control the timing and number of reflection samples obtained from the output of the photodetector 26 as the banknote 17 moves across the scanhead 18 and is scanned along its narrow dimension.

The detection of the border line is an important step that realizes improved discrimination efficiency since the border line serves as an absolute reference value at the start of scanning. If the edge of a banknote is used as a reference point, a relative shift of the scanning points could occur because of the random manner in which the distance from the edge to the border line changes from banknote to banknote due to the relatively high tolerance ranges allowed in printing and cutting banknotes. As a result, it becomes difficult to establish a direct correspondence between scanning points on successive banknote samples and the discrimination efficiency is adversely affected.

The use of the optical encoder to control the scanning process relative to the physical movement of a banknote across the scan head is also advantageous in that the encoder can be used to provide a predetermined delay following the detection of the border line prior to the scans. The encoder delay can be adjusted in such a way that the banknote is scanned only over those segments along its narrow dimension which contain the most distinguishable printed identifying feature relative to the different currency denominations.

For example, in the case of U.S. currency, it has been determined that the central, approximately two-inch area of the bills, when scanned across the central portion of the narrow dimension of the bill, provides sufficient data to distinguish between different U.S. currency denominations based on the correlation method of this invention. Similarly, the optical encoder may be used to to control the scanning process so that reflectance samples are taken for a predetermined period of time and are only taken after a certain period of time has elapsed when the border line has been detected, thereby limiting the scanning of the desired central region of the narrow dimension of the banknote.

The optical scanning and correlation method is based on using the above process to generate a series of characteristic specimen patterns or master patterns using "new" or "master" banknotes for each denomination of currency to be detected. According to a preferred embodiment, four characteristic patterns are generated and stored within a system memory, preferably in the form of an EEPROM 34 (see Fig. 1) for each currency denomination to be detected. The characteristic patterns for each banknote are generated respectively on optical scanning devices, i.e. with the process of obtaining the predetermined number of reflection samples carried out on the "top" and "bottom" side of the banknote, which is carried out along both the "forward" and "backward" directions relative to the patterns printed on the banknote.

For example, in adapting the invented method to U.S. currency, characteristic patterns are generated and stored for seven different denominations of U.S. currency, ie, $1, $2, $5, $10, $20, $50 and $100. Similarly, a set of 28 different characteristic patterns is stored within the system memory for subsequent correlation purposes. Once the characteristic sample patterns have been stored, the pattern generated by scanning a banknote during testing is processed by the CPU 30 compared with each of the 28 prestored characteristic sample patterns to produce for each comparison a correlation number representing the degree of correlation, ie, the similarity between corresponding ones of the plurality of data samples for the compared patterns.

The CPU 30 is programmed to identify the denomination of the scanned banknote as corresponding to the stored characteristic pattern for which the correlation number resulting from the pattern comparison is found to be the highest. To eliminate the possibility of misidentifying the denomination of a scanned banknote as well as to reduce the possibility of counterfeit banknotes being identified as corresponding to a valid denomination, a bi-level threshold of correlation is used as the basis for making a "positive" call, as will be explained in more detail below.

Using the above sampling and correlation methodology, the CPU 30 is programmed to count the number of banknotes belonging to a particular denomination of currency in a portion of a predetermined group of banknotes sampled for a predetermined sampling amount and to determine the total cumulative amount of currency represented by the banknotes sampled during a sampling batch. The CPU 30 is also connected to an output unit 36 adapted to provide an indication of the number of banknotes counted and the breakdown of the banknotes according to denomination of currency and the total amount of currency represented by the banknotes counted. The output unit 36 may also be adapted to provide a printout of the displayed information in a desired format.

Referring to Fig. 2, there is shown a block diagram representation of a preferred circuit arrangement for processing and correlating reflectance data in accordance with the system of this invention. As shown therein, the CPU 30 receives and processes a variety of input signals, including those from the optical encoder 32, the photodetector 26, and a memory unit 38, which may be a static random access memory (RAM) or an erasable programmable read only memory (EPROM). The memory unit 38 stores the correlation program based on which patterns are generated and test patterns are compared with pre-stored sample programs to identify the denomination of the test currency. A quartz 40 serves as a time base for the CPU 30, which is also provided with an external reference voltage VREF based on which peak detection of sensed reflectance data is carried out, as will be explained in more detail below.

The CPU 30 also receives a timer reset signal from a reset unit 44 which, as shown in Fig. 2A, receives the output voltage from the photodetector 26 and compares it by means of a threshold detector 44A relative to a predetermined voltage threshold, typically 5.0 volts, to provide a reset signal which goes "high" when a reflectance value corresponding to the presence of paper is sensed. More specifically, reflectance sensing is based on the assumption that no portion of the illuminated light strip (24 in Fig. 1) is reflected to the photodetector in the absence of a banknote positioned under the scanning head. Under these conditions, the output of the photodetector is a "dark" or "zero" level scan. The output of the photodetector changes to a reading of "white", which has typically been set to a value of about 5.0 volts, when the edge of the bill is first positioned beneath the scan head and falls below the light strip 24. When this occurs, the reset unit 44 provides a "high" signal to the CPU 30, marking the beginning of the scanning procedure.

In accordance with a feature of this invention, the thickness of the illuminated strip of light produced by the light sources within the scanning head is set to a relatively small value for the initial state of scanning when the thin border line is detected. The use of the narrow slit increases the sensitivity with which the reflected light is detected and allows for small changes in the "gray" level reflected from the surface of the banknote being sensed. This is important in ensuring that the thin border line of the pattern, i.e. the starting point of the printed pattern on the banknote, is accurately detected. Once the border line has been detected, subsequent reflection scanning is carried out based on a relatively wider strip of light to perform a complete scan across the narrow dimension of the banknote and to achieve the desired number of samples at a high speed. The use of a wider slit for the actual scanning also smoothes the output characteristics of the photodetector and realizes the relatively large value of the analog voltage, which is important for the accurate representation and processing of the detected reflection values.

Returning to Fig. 2, the CPU 30 processes the output of the photodetector 26 via a peak detector 46 which is essentially operative to sample the photodetector output voltage and capture the highest voltage value, i.e. the peak voltage value encountered after the detector allows it. The peak detector is also adapted to define a scaled voltage on the basis of which the pattern outline on the banknotes is detected. The detector 46 includes an ADC 48 for digitizing the photodetector output and includes a digital-to-analog converter (DAC) 50 for converting the signals back to an analog form based on a preselected reference voltage VREF from the voltage source 42. The output of the DAC 50 is fed through an inverting amplifier 42 to a voltage divider 54 which reduces the input voltage to a scaled voltage Vs representing a predetermined percentage of the peak value. The voltage Vs is based on the percentage of drop in the output voltage of the peak detector as it reflects the transition from the "high" reflectance value resulting from scanning the unprinted edge areas of a banknote to the relatively lower "gray" reflectance value resulting from encountering the thin border line. Preferably, the scaled voltage Vs is set at approximately 70 to 80% of the peak voltage.

The scaled voltage Vs is fed to a line detector 56 which is also provided with the incoming instantaneous output of the photodetector 26. The line detector 56 compares the two voltages on its input side and generates a signal LDET which normally remains "low" and goes "high" when the edge of the banknote is sensed. The signal LDET goes "low" when the incoming Photodetector output reaches the predefined percentage of peak photodetector output up to the point represented by voltage Vs. Thus, when signal LDET goes low, it is an indication that the border line of the banknote pattern has been detected. At this point, CPU 30 begins the actual reflection scanning under the control of encoder 32 (see Fig. 2), and the desired fixed number of reflection samples is obtained as the banknote moves across the illuminated light strip and is scanned along the central portion of its narrow dimension.

When generating the characteristic sample patterns, the reflection samples resulting from scanning a "new" banknote are entered into the appropriately designated sections within a system memory 60, which is preferably an EEPROM. The entering of the samples is carried out via a buffered address latch 58, if necessary. Preferably, the sample patterns are generated by scanning a "new" banknote a number of times, typically three times, and averaging the corresponding data samples before storing the average as representing a sample pattern. During currency discrimination, the reflectance values resulting from scanning a test banknote are sequentially compared under the control of the correlation program stored within the memory unit 38 with each of the corresponding characteristic patterns stored in the EEPROM 60, again via the address buffer 58.

Referring to Figs. 3 to 7, flow charts are shown illustrating the sequence of operations performed in 3 illustrates the sequence involved in sensing the presence of a bill under the scanhead and the border line on the bill. This section of the system program, designated "TRIGGER," is initiated at step 70. At step 71, a determination is made as to whether a start-of-note interrupt is set, which means that the system is ready to search for the presence of a bill, i.e., for it to occur. If the answer is found to be positive at step 71, step 72 is reached where the presence of the bill under the scanhead is determined based on the reset procedure as described above in connection with the reset unit 44 of FIG. 2.

If in step 72 the answer is found to be positive, i.e. that a banknote is present, step 73 is reached where a test is made to see if the border line has been detected based on the reduction in peak value for a predetermined percentage thereof, which is indicated as described above by the signal LDET going low. If in step 73 the answer is found to be negative, the program continues and loops until the border line has been detected. If in step 72 the answer is found to be negative, i.e. that no banknote is found to be present, the start-of-record interrupt in step 74 is reset and the program returns from the interrupt in step 75.

If the border line is found to have been detected in step 73, step 76 is reached where an A/D termination interrupt is enabled, indicating that analog-to-digital conversion can be subsequently performed at desired time intervals. Next, in step 77, the time when the first reflection sample is to be obtained is defined in conjunction with the output of the optical encoder. In step 78, the capture and digitization of the detected reflection samples is taken over by calling a routine designated "STARTA2D" which will be described in more detail below. At the completion of the digitization process, the end-of-note interrupt is enabled in step 79, which resets the system to detect the presence of the next banknote to be scanned. Subsequently, in step 80, the program returns from the interrupt. If the start-of-recording interrupt is not found to have occurred in step 71, a determination is made in step 81 to see if the end-of-recording interrupt has occurred. If the answer in step 81 is negative, the program returns from the interrupt in step 85. If an affirmative answer is obtained in step 81, step 83 is accessed where the start-of-recording interrupt is activated and in step 84 the reset unit monitoring the presence of a banknote is reset to be ready to determine the presence of banknotes. Subsequently, the program returns from the interrupt in step 85.

Referring to Figs. 4A and 4B, there are shown routines for starting the STARTA2D routine and for digitizing the routine itself, respectively. In Fig. 4A, initiating the STARTA2D routine in step 90 causes the sample pointer, which initializes a display of the sample obtained and digitized in a given time. Subsequently, in step 91, the special channel on which the analog-digital conversion is to be carried out is enabled. The interruption authorizing the digitization of the first sample and the main program is accessed again in step 93.

Fig. 4B is a flow chart illustrating the sequential procedure involved in the analog to digital conversion routine designated "A2D". The routine is started in step 100. Next, the sample indicator is decremented in step 101 to maintain an indication of the number of samples remaining to be obtained. In step 102, the digital data corresponding to the output of the photodetector for the current sample is read out. The data is converted to its final form in step 103 and stored within a predetermined memory segment as XIN.

Next, a check is made in step 105 to see when the desired fixed number of samples "N" has been taken. If the answer is found to be negative, step 106 is accessed where the interrupt enables digitization of the subsequent sample and the program returns from the interrupt in step 107 to complete the remainder of the digitization process. However, if the answer in step 105 is found to be positive, i.e., that the desired number of samples has been obtained, a flag indicating this is set in step 108 and the program returns from the interrupt in step 109.

Referring to Figure 5, there is shown the following procedure involved in executing the routine labeled "EXEC" which executes the mathematical steps involved in the correlation process. The routine is started in step 110. In step 111, all interrupts are disabled while the CPU program start occurs. In step 112, certain constants associated with the sampling process are set, and in step 113, communication protocols for exchanging processed data and associated results, baud rates, interrupt masks, etc. are defined.

At stage 114, the reset unit indicating the presence of a banknote is reset to detect the presence of the first banknote to be scanned. At stage 115, the start recording interrupt is enabled to set the system to search for the first incoming banknote. Subsequently, at stage 116, all other related interrupts are enabled, since at this point the initialization process has been completed and the system is ready to begin scanning the banknote. A check is made at stage 117 to see if all of the desired number of samples have in fact been obtained. If at stage 117 the answer is found to be negative, the program loops until a positive answer is obtained. At this time, stage 118 is accessed where a flag is set to indicate the initialization of the correlation procedure.

In accordance with this invention, a simple correlation procedure is used to process the digitized reflectance values into a form that can be compared with corresponding values pre-stored in an identical format are appropriate and accurate. More specifically, as a first step, the average value for the group of digitized reflection samples (cf. "n" samples) obtained for a scanning run for a banknote is first obtained as follows:

Subsequently, a normalization factor Sigma "" is determined which is equal to the sum of the squares of the difference between each sample and the mean when normalized by the total number of samples. In more detail, the normalization factor is calculated as follows:

In the final stage, each reflection sample is obtained by obtaining the difference between the sample and the average value calculated above and dividing it by the square root of the normalization factor Sigma "" as defined by the following equation:

The result of using the above correlation equations is that following the normalization process, there is a relationship of correlation between a test pattern and a sample pattern such that the total sum of the products of the corresponding samples in a test pattern and any sample pattern, when divided by the total number of the samples is equal to one if the patterns are identical. Otherwise, a value other than one is obtained. Accordingly, the correlation number or correlation factor resulting from the comparison of the normalized samples within a test pattern with those of a stored sample pattern provides a clear indication of the degree of similarity or correlation between the two patterns.

According to a preferred embodiment of this invention, the fixed number of reflectance samples digitized and normalized for a banknote scan is selected to be 64. It has been found experimentally that the use of higher binary orders of samples (such as 128, 256, etc.) does not provide a correspondingly higher discrimination efficiency in relation to the increased processing time involved in executing the correlation procedure described above. It has also been found that the use of a binary order of samples less than 64, such as 32, produces a significant drop in discrimination efficiency.

The correlation factor may be conveniently represented in binary terms to facilitate correlation. In a preferred embodiment, the unity factor resulting when a correlation of 100% exists is represented in terms of the binary number 2¹⁰, which is equal to a decimal value of 1024. Using the above procedure, the normalized samples within a test pattern are compared with those of the 28 sample patterns stored within the system memory to determine the particular stored patterns to which the test patterns are assigned by Identify the comparison that results in a correlation number close to 1024.

According to a feature of this invention, a two-level threshold of correlation is required to be satisfied before a particular call is made. More specifically, the correlation procedure is adapted to identify the two highest correlation numbers resulting from the comparison of the test pattern with one of the stored patterns. At this point, a minimum threshold of correlation is required to satisfy both of these correlation numbers. It has been found experimentally that a correlation number of about 800 serves as a good limiting threshold above which positive calls can be made with a high degree of certainty and below which the identification of a test pattern as corresponding to one of the stored patterns is uncertain. As a second threshold level, a minimum separation between the two highest correlation numbers is specified before a call is made. This ensures that a positive call is only made if a test pattern does not match more than one stored sample pattern in a given correlation range. Preferably, the minimum separation between correlation numbers is set to a number between 100-150.

Returning to Fig. 5, the correlation procedure is started at step 119 where a routine labeled "PROCESS" is accessed. The procedure involved in executing this routine is illustrated in Fig. 6A, which shows that the routine starts at step 130. At step 131, the mean X is calculated based on equation (1). At step 132, the sum of squares is calculated in accordance with calculated using equation (2). In step 133, the digitized values of the reflection samples, which are represented as integers, are converted to a floating point format for further processing.

At step 134 a check is made to see if all samples have been processed and if the answer is found to be positive the routine ends at step 135 and the main program is again accessed. If the answer is found to be negative at step 134 the routine returns to step 132 where the above calculations are repeated.

At the end of the PROCESS routine, the program returns to the EXEC routine in step 120, where the flag indicating that all digitized reflectance samples have been processed is reset. Next, a routine labeled "SIGCAL" is accessed in step 121. The procedure involved in executing this routine is shown in Figure 6B, which shows that the routine starts in step 140. In step 141, the square root of the sum of squares as calculated by the PROCESS routine is calculated in accordance with equation (2). In step 142, the floating point values calculated by the PROCESS routine are normalized in accordance with equation (3) using the values calculated in step 141. In step 143, a check is made to see if all digital samples have been processed. If the answer is found to be negative in step 143, the program returns to step 142 and the conversion continues until all samples have been processed. At this point the answer in step 143 is positive and the routine returns to the main program in step 144.

Returning to the flow chart of Figure 5, the next stage to be executed is stage 122, where a routine called "CORREL" is accessed. The procedure involved in executing this routine is shown in Figure 7, which shows that the routine starts at stage 150. In stage 151, the correlation results are initialized to zero, and in stage 152, the test pattern is compared to the first of the stored master patterns. In stage 153, the first call corresponding to the highest correlation number obtained up to that point is determined. In stage 154, the second call corresponding to the second highest correlation number obtained up to that point is determined. In stage 155, a check is made to see if the test pattern has been compared to all of the master patterns. If the answer is found to be negative, the routine returns to step 152 where the comparison procedure is repeated. If all of the sample patterns have been compared with the test patterns, step 155 gives a positive result and the routine returns to the main program at step 156.

Returning to Fig. 5, in step 123, a flag indicating that the correlation procedure is complete is reset and step 124 is accessed where a routine designated "SEROUT" is initiated. It should be noted that steps 118 and 123, which are directed to setting and resetting the TP2 flag, operate primarily to provide a measure of the processing time involved in the entire correlation procedure. These steps may be omitted if processing time is not monitored. The procedure involved in executing the SEROUT routine is illustrated in Fig. 8, which shows that the routine starts at step 160. In step 161, the monetary denomination corresponding to the first call is converted to ASCII format and displayed. In step 162, the correlation number corresponding to the first call is converted to ASCII format and displayed.

At stage 163, the currency denomination corresponding to the second call is converted to ASCII format and displayed. At stage 164, the correlation number corresponding to the second call is converted to ASCII format and displayed. The routine then returns to the main program. At this point in the main program, the correlation procedure is completed and a related count of the identified denominations can be performed with the associated results and also displayed along with the corresponding calls and correlation numbers. After this correlation and display procedure is completed, the system is ready to initiate the process of scanning the next incoming banknote.

It should be noted that in carrying out the optical scanning and correlation method of this invention, separate microprocessor units are provided for (i) carrying out the scanning and correlation process and (ii) controlling the general functions of the overall system. In such an operation, the general processing unit would preferably be used to display the identified denominations and associated count results. In this procedure, the SEROUT routine (Fig. 8) would only involve the transfer of the banknote denomination and call information from the scanning and correlation processing unit to the general processing unit for subsequent display.

Referring to Figures 9A-C, there are shown three test patterns generated for the forward scan of a $1 bill along its top, the reverse scan of a $2 bill along its top, and the forward scan of a $100 bill along its top. It should be noted that for the purposes of clarity, the test patterns in Figures 9A-C were generated using 128 reflection samples per bill scan, as opposed to the preferred use of only 64 samples. The significant difference that exists between the corresponding samples for these three test patterns indicates the high degree of certainty with which currency denominations can be retrieved using the optical scanning and correlation procedure mentioned above.

The optical scanning and correlation method described above allows the identification of pre-programmed currency denominations with a high degree of accuracy and is based on a relatively low processing time for digitizing the scanned reflectance values and comparing them to characteristic sample patterns. The method is used to scan banknotes, to normalize the scanned data and to generate sample patterns in such a way that banknote scans during operation have a direct relationship between compared scan points in areas of the banknotes having the most distinguishable printed identifying features. A relatively small number of reflectance samples are required to be able to adequately distinguish between different currency denominations.

A major advantage of this approach is that it does not require banknotes to be cut along their width dimensions scanned. Conventional systems have been forced to adopt the scanning approach on the width dimension in order to achieve the larger number of samples typically required for accurate denomination discrimination. Furthermore, reducing the number of samples reduces processing time to such an extent that additional comparisons can be made during the time available between scanning of successive banknotes. More particularly, as described above, it becomes possible to compare a test pattern with at least four stored characteristic sample patterns so that the system is able to identify money scanned in the "forward" or "backward" direction along the "top" or "bottom" of the banknotes.

Another advantage arising from the reduction in processing time realized by the present scanning and correlation scheme is that the response time required each time the transport of a banknote identified as "counterfeit," i.e., which does not match one of the stored characteristic sample patterns or includes diverting such a banknote to a separate stacker bin, is correspondingly reduced. Accordingly, the system can be suitably programmed to set a flag when a scanned pattern does not match one of the sample patterns. Identification of such a condition can be used to stop the banknote transport drive motor for the mechanism. Since the optical encoder is tied to the rotation of the drive motor, synchronization between pre-shutdown and post-shutdown states can be maintained. In dual processor implementation as discussed above, the information relating to the identification of a "counterfeit" banknote would be included in the information that would be forwarded to the general processing unit, which in turn would control the drive motor in an appropriate manner.

The correlation procedure and the accuracy with which a denomination is identified is directly related to the degree of correspondence between reflectance samples on the test pattern and corresponding samples on the stored master patterns. Thus, shrinkage of the "used" banknotes, which in turn causes a corresponding reduction in their narrow dimension, may produce a drop in the degree of correlation between such used banknotes of a given denomination and the corresponding master patterns. Banknotes that have suffered a high degree of use will exhibit such reduction in both the narrow and wide dimensions of the banknotes. While the scanning and correlation method of this invention remains relatively independent of such changes in the width dimension of banknotes, the reduction along the narrow dimension can effect correlation factors by realizing a relative shift in reflectance samples obtained when "shrunken" banknotes are transported over the scanhead.

To compensate or cancel the effect of narrow dimension contraction, the correlation method described above can be modified using a progressive shifting technique whereby a test pattern that does not correspond to one of the sample patch patterns is divided into predefined sections and samples from successive sections are progressively shifted and again compared with the stored patterns to identify the denomination. It has been experimentally determined that such progressive shifting effectively counteracts a sample shift resulting from the contraction of a banknote along its narrow dimension.

The effect of progressive shifting is best illustrated by the correlation patterns shown in Figures 10A-D. For the sake of clarity, the illustrated patterns were generated using 128 samples for each banknote scan compared to the preferred use of 64 samples. Figure 10A shows the correlation between a test pattern (represented by a thick line) and a corresponding specimen pattern (represented by a thin line). It is clear from Figure 10A that the degree of correlation between two patterns is relatively low, showing a correlation factor of 606.

The manner in which the correlation between these patterns is increased by using progressive shifting is best illustrated by considering the correlation at reference points labeled AE along the axis defining the number of samples. The effect on the correlation produced by "simple" progressive shifting is shown in Fig. 10B, which shows a "simple" shifting of the test patterns of Fig. 10A. This is accomplished by dividing the test patterns into two equal segments, each comprising 64 samples. The first segment is retained without any shifting, the second segment being shifted by a factor of one data sample. In this condition, it was found that the correlation factor at the reference points included in the shifted section, especially at point E: has been improved.

Fig. 10C shows the effect produced by "double" progressive shifting, where sections of the test patterns are shifted in three stages. This is achieved by dividing the overall pattern into three approximately equal sections. Section one is not shifted, section two is shifted by one data sample (as in Fig. 10B), and section three is shifted by a factor of two data samples. It can be seen that with the "double" shifting, the correlation factor at point E is further increased.

On a similar basis, Fig. 10D shows the effect on the correlation produced by "triple" progressive shifting, where the overall pattern is divided into four (4) approximately equal sized sections. Subsequently, section one is retained without any shift, section two is shifted by one data sample, section three is shifted by two data samples, and section four is shifted by three data samples. It can be seen that under these conditions the correlation factor at point E was again increased.

Fig. 10E shows the effect on correlation produced by "four-way" shifting, where the pattern is first divided into five (5) approximately equal-sized sections. The first four (4) sections are shifted in accordance with the "three-way" shifting procedure of Fig. 10D, with the fifth section being shifted by a factor of four (4) data samples. It is clear from Fig. 10E that the correlation at point E is increased almost to the point of overlap of the compared data samples.

The advantage of using the progressive shifting technique over merely shifting a given amount of data samples across the entire test pattern is that the improvement in correlation achieved at the initial portions of the pattern as a result of the shifting is not cancelled out or offset by subsequent shifts in the test pattern. From the above figures, it can be seen that the degree of correlation for sample points falling within the progressively shifted portions is increased accordingly.

More importantly, progressive shifting realizes substantial increases in the overall correlation factor resulting from pattern matching. For example, the original correlation factor of 606 (Fig. 10A) is increased to 681 by the "simple" shifting shown in Fig. 10B. The "double" shifting shown in Fig. 10C increases the correlation number to 793, the "triple" shifting of Fig. 10D increases the correlation number to 906 and finally the "quadruple" shifting shown in Fig. 10E increases the total correlation number to 960. It has been found that using the above procedure, used banknotes which have a high degree of contraction in the narrow dimension and which cannot be accurately identified as belonging to the exact currency denomination if the correlation is carried out without shifting can be identified with a high degree of certainty using the progressive shifting procedure, preferably by using the "triple" or "quadruple" shifting.

Referring to Fig. 11, a money discrimination and counting device 210 is shown which embodies the principles of the present invention. The device comprises a housing 212 including left and right side walls 214 and 216, a rear wall 218 and a top generally designated 220. The device has a front portion 220 including a generally vertical front portion 224 and a forwardly inclined portion 225 including side portions provided with control panels 226A and 226B on which various control switches for operating the device and associated display devices are mounted.

For receiving a stack of banknotes 228 which are to be distinguished according to their denomination, an input container 227 is defined on the upper side 220 by a downwardly inclined support surface 229 on which a pair of vertically arranged side walls 230, 232 are provided which are connected by a vertically arranged front wall 234. The walls 230, 232 and 234 in combination with the inclined surface 229 define a housing where the stack of banknotes 228 is positioned.

From the input container, banknotes are moved along a three-part transport path including an input path where banknotes are moved along a first direction in a substantially planar position, a curved guide path where banknotes are received from the input path and guided in such a way that they change the direction of transport to a second different direction, and an output path where the banknotes are moved in a planar position along the second different direction via a money discriminator arranged downstream of the curved guide path, as will be described in more detail below. In accordance with With the improved optical scanning and correlation method according to this invention, the transport path is defined in such a manner that banknotes are picked up, transported along the input path, the curved guide path and the output path, and stacked with the narrow dimension "W" of the banknotes being maintained parallel to the transport path and the direction of movement at all times.

The forwardly inclined portion 225 of the document processing apparatus 210 includes a platform surface 235 centrally disposed between the side walls 214, 216 and is adapted to receive banknotes processed by the currency discriminator for delivery to a stacker plate 242 where the processed banknotes are stacked for subsequent removal. More particularly, the platform 235 includes an angled surface 236 connected thereto and is provided with openings 237, 237A from which resilient blades 238A, 240A of a corresponding pair of stacker wheels 238, 240 extend outwardly, respectively. The stacker wheels are supported for rotational movement about a stacker shaft 241 disposed about the angled surface 236 and suspended from the side walls 214 and 216. The resilient blades 238A, 240A of the stacker wheels cooperate with the stacker platform 235 and the openings 237, 237A to receive banknotes delivered thereto. The blades are operative to sequentially deliver such banknotes to the stacker plate 242 which is connected to the angled surface 236 and which also receives the stacker wheel openings and the wheels projecting therefrom. During operation, a banknote delivered to the stacker platform 235 is received by the resilient blades and is received between a pair of adjacent blades which in combination form a curved housing which slows entry of a banknote thereinto and serves as a means for supporting and transferring the banknote from the stacker platform 235 to the stacker plate 242 as the stacker wheels rotate. The mechanical configuration of the stacker wheels and resilient blades provided thereon, as well as the manner in which they cooperate with the stacker platform of the stacker plate, is conventionally known and accordingly will not be described in detail herein.

The banknote processing and counting apparatus 210 is provided with means for receiving and "stripping" individual banknotes from banknotes stacked in the input tray 237. To provide this stripping action, a feed roller 246 is rotatably suspended about a drive shaft 247 which in turn is supported by the side walls 214, 216. The feed roller 246 projects through a slot provided on the downwardly inclined surface 229 of the input tray 227 which defines the input path and is in the form of an eccentric roller, at least a portion of the periphery of which is provided with a relatively high friction producing surface 246A. Surface 246A is adapted to engage the bottom bill of bill stack 228 as roller 246 rotates, initiating advancement of the bottom bill along the feed direction represented by arrow 247B (see Figure 13). Specifically, the eccentric surface of feed roller 246 "jiggles" the bill stack once per revolution so as to loosen and loosen the bottom bill within the stack, thereby facilitating advancement of the bottom bill along the feed direction.

The action of the feed roller 246 is supplemented by the provision of a capstan or drum 248 suspended for rotation about a capstan drive shaft 249 which in turn is supported by the side walls 214 and 216. Preferably, the capstan 248 includes a centrally located friction roller 248A having a smooth surface and formed of a friction producing material such as rubber or hard plastic. The friction roller is sandwiched between a pair of capstans 248B and 248C with at least a portion of the outer periphery provided with a high friction producing surface 248D.

The friction surface 248D is very similar to the friction surface 246A provided on the feed roller and allows the capstan roller to frictionally advance the lower bill along the feed direction. Preferably, the rotational movement of the capstan roller 248 and the feed roller 246 is synchronized in such a way that the friction surfaces provided on the peripheries of the capstan roller and the feed roller rotate in unison, thereby causing a matching frictional contact between the lower bill of the bill stack 228.

To ensure effective contact between the drive drum 248 and a banknote being jolted by the feed roller 246 and in the process of being advanced, picker rollers 252A, 252B are provided for exerting a uniform downward force on the leading edge of the banknotes resting in the input tray 227. The picker rollers are supported on respective picker arms 254A, 254B, which in turn are supported for circular arc motion about a support shaft 256 extending over the side walls of the apparatus. The sorting rollers are freely rotatable about the sorting arms and when no banknotes are in contact with the drive drum 248, they will rest on the friction roller 248A and will accordingly counter-rotate therewith. However, when banknotes are present and in contact with the drive drum 248, the sorting rollers will rest on the leading edges of the banknotes and exert a direct downward force on the banknotes since the rotation of the rollers is inhibited. The result is that the advancing action exerted by the contact between the friction-inducing surfaces 248D on the drive drum rollers 248B, 248C is increased, thereby facilitating the stripping of a single banknote one at a time from the stack of banknotes 228.

Between the sorting arms 254A, 254B, the support shaft 256 also supports a separator arm 260 which carries at its end remote from the shaft a stationary stripper shoe 258 provided with a friction surface which imparts a frictional pull to the banknotes on which the sorting rollers rest. The separator arm is mounted for rotation about the support shaft 256 and is spring loaded in such a manner that it rests on the drive drum with a selected amount of force.

In operation, the sorting rollers rotate with the rotation of the friction roller 248A due to its freely rotating nature until they encounter the leading edges of one or more banknotes. At this point, the rotation of the sorting rollers stops and the leading edges of the banknotes are forced into positive contact with the friction-inducing surfaces on the periphery of the drive drum rollers. The effect is to to push the lower banknote away from the rest of the banknotes along the direction of rotation of the drive drum. At the same time, the separator shoe 258 also rests on one of the banknotes which are driven forward by the drive drum rollers.

The tension on the sorting arm 254A is selected such that the downward force exerted on such driven bill only allows a single bill to be advanced. If the event occurs that two or more bills are driven out of the contact made by the sorting rollers and the drive drum rollers, the downward force exerted by the spring loaded shoe should be sufficient to inhibit further forward movement of the bills. The tension under which the sorting arm is spring loaded can be suitably adjusted to control the downward force exerted by the shoe in such a manner as to coincide with the bill stripping action produced by the sorting rollers and the drive drum rollers. Thus, the possibility of more than two banknotes being driven forward at the same time due to the rotation of the drive shaft is considerably reduced.

The banknote transport path includes a curved guide path 270 provided in front of the drive shaft 248 for receiving banknotes that have been driven forward along the input path defined by the front portion of the inclined surface 229 into frictional contact with the rotating drive drum. The guide path 270 includes a curved portion 272 substantially corresponding to the curved periphery of the drive shaft 248 to match the impetus provided by the drive drum rollers 248B, 248C on a withdrawn banknote.

A pair of idler rollers 262A, 262B are provided downstream of the sorting rollers for guiding the bills driven by the drive drum 248 into the curved guideway 270. More specifically, the idler rollers are mounted on respective idler roller arms 264A, 264B which are mounted for circular arc movement about a idler roller shaft 266 which in turn is supported by the side walls of the apparatus. The idler roller arms are spring loaded on the idler roller shaft so that a selected downward force can be exerted by the idler rollers on a withdrawn bill, thereby ensuring continued contact between the bill and the drive roller 248 until the bill is guided into the curved portion 270 of the guideway 270.

Downstream of the curved portion 272, the banknote transport path has an output path for the banknotes. The output path is provided in the form of a flat portion 274 along which banknotes guided along the curved guide path 270 by the rollers 262A, 262B are moved along a direction opposite to the direction along which banknotes are moved out of the input container. The movement of the banknotes along the direction of rotation of the drive drum caused by the sorting rollers 252, 252B and the drive rollers 248B, 248C and the guidance provided by the portion 272 of the curved guide track 270 changes the direction of movement of the banknotes from the initial movement along the inclined surface 229 of the input tray 227 (see arrow 247B in Fig. 13) to a direction along the flat portion 274 of the output track, as best shown in Fig. 13 by the arrow 272B.

Thus, a bill stripped from the stack of bills in the input tray is initially moved along the input path under the positive contact between the sorting rollers 252A, 252B and the drive drum rollers 248B, 248C. The bill is subsequently guided by the curved guide path 270 under the positive contact with the idler rollers 262A, 262B onto the flat portion 274 of the output path.

In the output path, banknotes are positively guided along the planar portion 274 by a transport roller assembly including a plurality of axially spaced, positively driven transport rollers 282A, 284A, 286A mounted on a transport shaft 287 supported by the side walls of the apparatus. The planar portion is provided with apertures through which at least two of the transport rollers, specifically rollers 282A and 284A, project into counter-rotating contact with corresponding free-running passive rollers 292A, 294A. The passive rollers are mounted on a support shaft 295 supported between the side walls of the apparatus beneath the planar portion 274 of the output path. The passive transport rollers 292A, 294A are spring loaded into counter-rotating contact with the active transport rollers 282A, 284A, 286A, and the points of contact are coplanar with the output path so that banknotes can be moved along the path in a planar manner under positive contact of the opposed active and passive rollers. A similar group of active transport rollers 282B, 284B, 286B and opposed spring loaded passive transport rollers 292B, 294B are provided downstream of the first group of transport rollers at a distance just as short as the length of the narrow Dimension of the banknotes to be distinguished. Furthermore, the distance between the idler rollers 262A, 262B and the first group of transport rollers is selected in such a way that a banknote guided along the curved guide track 259 is drawn into contact between the first group of active and passive transport rollers before the banknote moves away from the forced contact between the idler rollers 262A, 262B and the drive drum 248.

The active transport rollers are driven at a speed that is significantly higher than that of the drive drum rollers. Since the passive rollers are freely rotating and the active rollers are positively driven, the first group of transport rollers causes a banknote that comes between the rollers along the flat portion of the output path to be drawn into a nip formed between the active and passive rollers. The higher speed of the active transport rollers imparts an abrupt acceleration to the banknote, which acceleration functions to separate or strip the banknote from other banknotes that may be guided into the curved guide path together with the banknote that is under the action of the transport rollers.

Downstream of the first group of transport rollers, banknotes are moved along the flat section into the nip formed between the second group of active and passive transport rollers, which are driven at the same speed as the first group of transport rollers. Preferably, the opposing groups of transport rollers 282A to 282B, 284A to 284B and 286A to 286B are coupled together by a belt 290 so that the positive rotational action of the transport shaft 287 is imparted to the rollers. which are carried on the second transport shafts 288. The arrangement of the second group of transport rollers is such that the positive contact exerted by the rollers on a banknote moving along the output path occurs before the banknote is released from the positive contact between the first group of transport rollers. The second group of transport rollers thus positively feeds a banknote onto the stacker platform 235 from where the stacker wheels 238, 240 pick up the banknote and deposit it on the stacker plate 242.

Referring particularly to Figures 14 and 15, there are shown side and top views, respectively, of the document processing apparatus of Figures 11 to 13, illustrating the mechanical arrangement for driving the various means for transporting the banknotes along the three sections of the transport path, i.e., along the input path, the curved guide path and the output path. As shown therein, a motor 300 is used to impart rotary motion to the drive drum shaft by means of a belt-pulley arrangement comprising a pulley 310 provided on the drive drum shaft 249 and connected to a pulley 304 provided on the motor drive shaft by a belt 306. The diameter of the drive pulley 310 is selected to be suitably larger than the motor pulley 304 to achieve the desired speed reduction from the typically high speed at which the motor 300 operates.

The drive shaft 247 for the drive roller 246 is supplied with a rotary motion by means of a pulley 308 provided on the drive shaft and which is connected via a belt 310 to a corresponding pulley 312 which on the drive roller shaft 249. The pulleys 308 and 310 have the same diameter so that the drive roller shaft 247 and hence the drive roller 246 rotate in unison with the drive drum 248 which is mounted on the drive drum shaft 249.

To impart rotational movement to the transport rollers, a pulley 314 is mounted on the transport roller shaft 287 corresponding to the first group of transport rollers and is connected to a corresponding pulley 316 on the drive drum shaft 249 by a belt 318. The diameter of the transport roller pulley 314 is selected to be suitably smaller than the diameter of the corresponding drive drum pulley 316 so as to realize an increase in the speed from the drive drum rollers to the transport rollers. The second group of transport rollers mounted on the transport roller shaft 288 is driven at the same speed as the rollers of the first group of transport rollers by means of a pulley 320 which is driven by means of a pulley 320 which is connected to the transport pulley 314 by means of a belt 322.

As also shown in Figures 14 and 15, an optical encoder 299 is mounted on one of the transport roller shafts, preferably on the passively driven transport shaft 288, to precisely follow the transverse displacement of the banknotes supported by the transport rollers with respect to the rotational movement of the transport shafts, as explained in detail above in connection with the optical scanning and correlation method of this invention.

To drive the stacker wheels 238, 240, an intermediate pulley 322 is mounted on a suitable support (not shown) and is connected to a corresponding pulley 324 provided on the drive roller shaft 249 by a belt 326. Because of the time required to transport the banknotes stripped from the stack of currency in the input tray through the three-part transport path and onto the stacker platform, the speed at which the stacker wheels can rotate to deliver processed banknotes onto the stacker plate is necessarily less than that of the drive drum shaft. Accordingly, the diameter of the intermediate pulley 322 is selected to be larger than the diameter of the corresponding drive drum pulley 324 so as to realize a reduction in speed. The intermediate pulley 322 has an associated pulley 328 which is connected to a stacker pulley 330 provided on the drive shaft 241 for the stacker wheels 238, 240 by means of a belt 332. In the preferred embodiment shown in Figures 11 to 15, the stacker wheels 238, 240 rotate in the same direction as the drive drum rollers. This is accomplished by placing the belt 332 between the pulleys 328, 330 in a "figure 8" configuration around an anchor pin 333 located between the two pulleys.

The curved portion 272 of the guideway 270 is provided on its underside with an optical sensor arrangement 299 which includes an LED 298 for carrying out standard money processing operations such as counterfeit detection using conventional methods, doubles detection, length detection, skew detection, etc. In addition to However, in conventional arrangements, money discrimination according to face value is not carried out in this area, for reasons described below.

According to a feature of this invention, optical scanning of the banknotes is carried out in accordance with the improved optical scanning and correlation method described above by means of an optical scanhead 296 disposed downstream of the curved guideway 270 along the planar portion 274 of the output path. More specifically, the scanhead 296 is disposed beneath the planar portion of the output path between the two sets of transport rollers. The advantage of this methodology is that optical scanning is carried out on banknotes when they are held in a substantially planar position as a result of the positive contact between the two sets of transport rollers at both ends of the banknote along its narrow dimension.

It should be understood that the drive arrangement described above is for illustrative purposes only. Alternative arrangements for imparting the necessary rotational motion to produce movement of the banknotes along the three-part transport path may be used as is most effective. It is important, however, that the surface speed of the banknotes over the two groups of transport rollers is greater than the surface speed of the banknotes over the drive drum rollers in order to achieve optimum banknote separation. It is this difference in speed that produces the abrupt acceleration of the banknotes as the banknotes come into contact with the first group of transport rollers.

The drive assembly may also include a one-way clutch (not shown) provided on the drive drum shaft, and the drive drum shafts, the transport roller shafts and the stacker wheel shafts may be fitted with flywheel assemblies (not shown). The combination of the one-way clutch and flywheels may be advantageously used in expediting batch processing of banknotes by ensuring that any banknotes remaining in the transport path after currency discrimination are automatically drawn away from the transport path into the stacker plate as a result of the inertial dynamics of the flywheel assemblies.

As described above, the operation of the optical scanning and correlation method of this invention requires only a relatively small number of reflection samples to adequately discriminate between different currency denominations. Thus, highly accurate discrimination is possible even when banknotes are scanned along their narrow dimension. However, the accuracy with which a denomination is identified is based on the degree of correlation between the reflection samples on the test pattern and corresponding samples on the stored sample patterns. Accordingly, it is important that banknotes be transported over the discriminator in a level position, and more importantly, at a uniform speed.

This is achieved in the banknote processing apparatus of Figs. 11 to 15 by positioning the optical scanning head 296 on one side of the planar portion 274 of the output path between the two groups of transport rollers. In this area, banknotes are held in positive contact with the two groups of rollers, thereby ensuring that the banknotes move across the scanning head in a substantially planar manner. Furthermore, a uniform speed of banknote movement is maintained in this region since the second group of passive transport rollers is driven at a speed identical to that of the active transport rollers by means of the belt connecting the two groups of rollers. Locating the optical scanning head 296 in such a manner downstream of the curved guideway 270 along the planar portion 274 maintains a direct correspondence between reflection samples obtained by optically scanning the banknotes to be discriminated and the corresponding samples in the stored sample patterns.

According to a preferred embodiment, the optical scanhead includes a plurality of light sources that act in combination to uniformly illuminate stripes of light of the desired dimension on banknotes positioned on the transport path beneath the scanhead. As shown in Figure 16, the scanhead 296 includes a pair of LEDs 340, 342 that direct light beams 340A and 340B, respectively, downwardly toward the planar portion 274 of the output path against which the scanhead is positioned. The LEDs 340, 342 are angularly arranged relative to the vertical axis Y in such a manner that their respective light beams combine to illuminate the desired stripe of light 342.

The scanning head 296 includes a photodetector 346 centrally located directly above the strip for sensing the light reflected off the strip. The photodetector 346 is connected to a central processing unit (CPU) (not shown) for processing the scanned data in accordance with the principles of this Invention. Preferably, the non-radiating rays 340A, 340B from the respective LEDs 340, 342 pass through an optical mask 343 to realize the illuminated strips of the desired dimensions.

In order to collect the reflectance samples with high accuracy, it is important that the photodetector collect reflectance data uniformly across the illuminated strip. In other words, if the photodetector 346 is positioned centrally above the light strip relative to its center "0", the output of the photodetector as a function of the distance from the center "0" along the X-axis should optimally approximate a step function as shown by curve A in Fig. 17. With the use of a single light source angularly offset relative to the vertical, the change in the photodetector output typically approximates a Gaussian function as shown by curve B in Fig. 17.

In accordance with a preferred embodiment, the two LEDs 340 and 342 are angularly arranged relative to the vertical axis by angles α and β, respectively. The angles α and β are selected such that the resulting output of the photodetector is as close as possible to the optimal distribution curve A in Figure 17. According to a preferred embodiment, the angles α and β are each selected to be 19.9°. The photodetector output distribution realized by this arrangement is represented by the curve labeled "C" in Figure 17, which effectively combines the individual Gaussian distributions of the light sources to achieve an overall distribution that is suitably approximated to the optimal curve A.

The manner in which the plurality of light stripes of different dimensions are produced by the optical scanning head by means of an optical mask is illustrated in Figure 18. As shown therein, the optical mask 350 specifically includes a generally opaque region 352 on which two slits 354 and 356 are defined to allow light from the light sources to pass therethrough so as to illuminate light stripes of the desired dimensions. More specifically, the slit 354 corresponds to the wide stripe used to obtain the reflection samples corresponding to the characteristic pattern for a test bill. According to an illustrative embodiment, the wide slit has a length of about 0.300 inches and a width of about 0.050 inches. The second slot 356 is adapted to create a relatively narrow illuminated strip used to detect the thin border line surrounding the printed identification feature on banknotes, as described in detail above. According to the illustrative embodiment, the narrow slot 356 has a length of approximately 0.300 inches and a width of approximately 0.010 inches.

It is obvious that a high machining accuracy would be required to precisely define the slots. In practice, it becomes difficult to machine the narrow strip 356 on the optical mask 350. This problem is approached according to a preferred embodiment by defining the mask 350 in the form of separate sections 360 and 362. The section 360 has one edge machined to correspond to one half section 356A of the desired slot 356. The second section 362 has a corresponding edge machined to correspond to the other half 356B of the slot 356. When the two sections 360 and 362 are mechanically joined together, they effectively define the narrow strip 356. The advantage with this approach is that both halves 356A, 356B which together define the strip 356 can be precisely defined, since machining on the edges of the mask can be handled with greater precision than machining within the mask itself.

Claims (10)

1. A money counting and evaluating device for receiving a stack of banknotes, rapidly counting and evaluating all of the banknotes in the stack, and then restacking the banknotes, the device comprising: a feeding mechanism (12, 14; 227, 246, 248) for receiving a stack of banknotes (17; 228) and for individually feeding the banknotes to a banknote transport mechanism (16; 282-290); wherein the banknote transport mechanism (16; 282-290) is arranged to transport banknotes from the feed mechanism (12, 14; 227, 246, 248) to a stacking station (20; 238, 242, 246); an optical scanning head (18; 296) disposed between the feed mechanism (12, 14; 227, 246, 248) and the stacking station (20; 238, 242, 246) for scanning a preselected segment of a region of each banknote transported by the transport mechanism (16; 282-290), the scanning head (18; 296) including at least one light source (22; 340, 342) for illuminating a strip of the preselected segment of a banknote and including at least one detector (26; 346) for receiving light from the illuminated strip on the banknote and for generating an output signal representing the changes in the intensity of the received light; means (30) for sampling the output signal at preselected intervals when a banknote is moved over the scanning head (18; 296); a memory (34) for storing characteristic signal samples generated by scanning the preselected segments of the banknotes of different denominations with the scanning head (18; 296) and sampling the output signal at preselected intervals; and a signal processing device (30) for receiving the signal samples and (A) determining the denomination of each scanned banknote (17) by comparing the stored signal samples with the output signal samples generated by scanning each banknote with the scanning head (18; 296), and (B) counting the number of scanned banknotes of each denomination, characterized in that the optical scanning head (18; 296) is provided for scanning a preselected segment of a central region of each banknote; the means (30) for sampling the output signal provides the output signal samples, the sample value being proportional to the intensity of light received from a different strip of the preselected segment of a banknote; wherein the memory (34) is provided for storing signal samples which are proportional to the intensity of the light received from different stripes of the preselected segment of a banknote; wherein the signal processing device (30) is arranged for comparing the stored signal samples which are proportional to the intensity of the light with the output signal samples which are proportional to the intensity of the light generated by scanning each banknote with the scanning head (18; 296); and wherein the signal processing means (30) is further arranged for (C) summing the cumulative value of the scanned banknotes of each denomination. 2. A currency counting and evaluating apparatus as claimed in claim 1, wherein: the bill transport mechanism (16) transports bills in the direction of the narrow dimension (W) of the bills; wherein: the optical scanning head (18) is stationary; wherein at least one detector (26) receives reflected light from the illuminated stripe on the bill and produces an output signal representing the changes in the intensity of the reflected light; and wherein each of the stored signal samples is proportional to the intensity of the light reflected from a different stripe of the preselected segment of a bill. 3. A money counting and evaluating apparatus according to claim 1 or 2, including a code converter (32) coupled to a transport mechanism (16) for monitoring the movement of each banknote by generating a repeating sequence signal synchronized with incremental movements of the transport mechanism, and including means for placing each banknote in a narrow Engagement with the transport mechanism and the scanning head (18) to ensure a fixed relationship between increments of movement of each banknote and corresponding increments of movement of the transport mechanism synchronized with the encoder. 4. A money counting and evaluating device according to claim 3, including means for sampling the output signal at increments synchronized with the repetitive tracking signal and at the same increments used in the characteristic signals stored in the memory (34). 5. A money counting and evaluating device according to any one of claims 1 to 4, wherein the detector (26) in the scanning head (18) is a single photodetector (346) which produces an electrical output signal proportional to the intensity of the light reflected from the scanned banknote. 6. A money counting and evaluating apparatus according to any one of claims 1 to 5, further comprising means for detecting a border line (17A) around the image printed on each banknote, and wherein the preselected segment is located within the border line (17A), and wherein the scanning of the preselected segment is initiated at a prescribed interval following the detection of the border line (17A). 7. A money counting and evaluating device according to any one of claims 1 to 6, wherein the preselected segment of each banknote is located in the central region of the banknote. 8. A currency counting and evaluating device according to any one of claims 1 to 7, wherein the feed and stacking stations (227, 246, 248; 238, 242, 246) are both located at the front of the device, and wherein the transport mechanism (282-290) carries the banknotes backwards away from the feed station and then returns the banknotes forwards to the stacking station (238, 242, 246). 9. A money counting and evaluating device according to claim 8, wherein the transport mechanism (282) forms a linear path for the banknotes on the upstream side of the stacking station (20; 238, 242, 246), the scanning head (18; 296) being arranged along the linear path. 10. A money counting and evaluating device according to any one of claims 1 to 9, wherein the light source (22; 340, 342) illuminates the preselected segment of each banknote from opposite sides of the detector (26; 348).