EP1421999A2 - Process for identification, classification and sorting of objects and materials and according recognition system - Google Patents

Abstract

Method for identification, classification and sorting of objects, and materials in which in a first method step, spectral data, e.g. color, chemical properties and spatial data, e.g. shape, dimensions, position and structure are determined for an object, objects or material moving relative to a measurement unit. The measurements are made in two dimensional planes with high spatial resolution and complete spectral resolution in real time. The measured data are used for real-time classification and sorting. The invention also relates to a corresponding recognition system.

Description

The invention relates to a method for identifying, classifying and sorting objects, objects and materials Detection system and an arrangement for performing this method.

Identification processes are used in a number of industrial processes or classification of objects, objects and materials an important instrument for process management and control.

Thus, according to DE-Al 19751862, an automated sorting process specified at which the identification of the to be sorted Objects is performed using spectral data analysis. ever after activation of the scanning device upstream of the spectrometer during the identification process, the spectral data for a predetermined amount of objects to be sorted and determined for the subsequent classification is forwarded to a computer unit. This process is complex in terms of apparatus technology, because such a recognition system from three subsystems, namely an image analysis system, a scanner and a NIR (near infrared) spectrometer exists.

Another well-known method for classifying objects, Objects and materials are the measurement of spectral properties on a measurement object in one or more optical wavelength ranges and the evaluation of the spectral absorption, Emission or reflection patterns. Set these procedures however, the measurement spot must be representative of the whole Target is. If this is not guaranteed, for example if there is a possibility that the object is made of other materials is contaminated or covered, is inhomogeneous, or if the Measurement objects have a relevant surface structure that is captured these methods are inadequate. For such Cases, the use of systems is required with which records the required spectral information in a spatially resolved manner can be.

It is known to use filter systems in this area optionally discrete filters when observing fewer, more selected Wavelengths are used or where by using tunable optical filters a multispectral image, in Combination with a recording unit, usually a camera, is recorded. Such a system using acoustic tunable optical filters are described in US-A-5,216,484. The limitation when using filter systems with more as an observed wavelength is that the measurement object not relative to the measuring device during the duration of the measurement allowed to move. When a relative movement occurs, a series receive images that include individual images of the measurement object individual wavelengths, but in different positions, whereby a complete image of a moving measurement object is not obtained can be. Due to this limitation, the requirement of Measurement with high resolution in both flat dimensions and in Real time without interrupting the process of such systems not be met.

The object of the present invention is to overcome the disadvantages of Prior art methods for identifying and Classifying objects, objects and materials, while at the same time ensuring a simple procedure should be.

According to the invention, a method of the type mentioned at the beginning is proposed, which is characterized in that in one process step the spectral data, such as color information, chemical Properties and spatial data such as shape, size, location and / or Structure of moving or moving relative to the measuring unit Objects, objects and materials in both flat dimensions with high spatial and full spectral resolution in real time are determined and recorded.

Advantageous embodiments of this method are the subject of Dependent claims.

The spectral ranges are preferably the following ranges and / or sub-areas of the areas and / or combinations of the areas for use: ultraviolet (UV, 200 - 380 nm), visible light (380 - 780 nm), near infrared (NIR, 780 - 2500 nm) and medium Infrared (> 2.5 µm). The objects, objects and Materials can be classified and classified based on this data in real time then sorted automatically, for example.

An exemplary application for the method according to the invention is the material or sorting of objects, Objects and materials, such as waste materials, but also processing of recycling materials. It is known to be of all kinds Items such as waste and recycled material by hand sort by. It is based on criteria such as color, shape and composition of the objects to be sorted. Although one Manual pre-sorting, such as with smaller amounts of waste is carried out in households, a relief for the further waste disposal is still manual sorting complex to carry out. To simplify the sorting or the Sorting of health-endangering or unhealthy The method according to the invention can enable substances here be used.

Another application example is the use of the invention Process for in-line quality control of products during the production process, for example surface coatings. The use of the method according to the invention enables distribution, homogeneity control for this application and quality of a coating process in real time. additionally can be achieved by using image processing algorithms automated on the classification result and real-time statements, for example via the spatial and / or the size distribution of irregularities hit what as a rule and / or quality assurance parameter for the process control of essential Is important for efficient and trouble-free production. Compared to the state of the art, namely (i) off-line sample tests, (ii) non-spatially resolved on-line cumulative test methods, in the application example surface resistance and / or capacitance, or (iii) the, from a technical and financial point of view complex use of several, across the width of the conveyor distributed individual measuring heads, represents the invention The process is therefore a significant step forward.

The invention further relates to a detection system consisting of a detection unit for the simultaneous determination of the spatial resolved image and spectral data in real time, which is above the distributed on a transport system, preferably a conveyor track, positioned to classify and / or sort objects and a detection and identification connected to this unit Evaluation unit for evaluating the determined multispectral Image data. Alternatively, especially when considering Surfaces of large and / or heavy objects, the detection unit moved on or by a single or multi-axis movement system with the test object in the rest position.

In both arrangements, lines can cross the direction of the relative Movement of the objects to be classified through a gap in the Detection system are mapped. Advantageously, there is this detection system from an imaging spectrograph in Combination with a recording unit. This system is through it marked that lines which are transverse to the direction of relative movement run between measurement object and detection unit, with Mapped using a suitable, spectrally coordinated optics and fed to an imaging spectrograph via an entry slit with a wavelength-dispersive element is equipped for spectral decomposition of the lines, so that at the Exit side of the spectrograph the incoming light into the corresponding spectral ranges and disassembled on a recording unit is mapped.

The data determined as explained above are subsequently passed on to the evaluation unit, which advantageously consists of an electronic data processing system. The one with high Speed transmitted data from the detection unit first with the help of an appropriate computer software program read. The computer software includes for further processing a set of mathematical algorithms to work with Determined with the help of the recognition system / systems Data streams in real time.

Due to the relative movement, lines are recorded sequentially and using computer software to create a multidimensional one Information space composed. This consists of the two Location coordinates in the longitudinal and transverse directions to the direction of movement as well as from a spectral dimension, which the spectral information describes as spectral intensity depending on the wavelength. By evaluating this data, it is possible to have both shape and size and location of the objects to be classified as well as simultaneously Color and / or other material properties, e.g. the chemical Composition to determine.

In general, this classification serves to assign the recognized ones Objects in groups with similar or comparable properties, as in the case according to the invention, shape, size, color, structure and / or chemical constitution, as well as combinations of these and / or similar parameters. Colors can also be used Hold on with the help of common color cameras, but these are on three color channels limited, so a color classification only with high technical effort or reduced precision can. However, the method according to the invention enables spectral decomposition enables, which is typical for 100 to 200 wavelengths is so that it can be precisely identified and classified the objects to be sorted is guaranteed.

By restricting the spectral range that can be recorded with one unit it may be advantageous for specific applications that Complete system with two or more detection units, each consisting from spectrally coordinated optics, imaging spectrograph and camera to equip. This makes it possible to base the information broaden the classification and thus the reliability of the classification to increase. The detection units become this preferably integrated into the overall system so that they are synchronous run towards each other and capture the same observation field. The Detection units can either be shared or individual Detection and evaluation units are connected.

The invention further relates to an arrangement for performing the inventive method as an automated material sorting process, which essentially consists of a continuous conveyor track and a device that is actually connected to the detection system, preferably a row of nozzles. Advantageous configurations this arrangement is disclosed according to subclaims.

Fig. 1 eine mögliche Vorrichtung zur Durchführung des erfindungsgemäßen Verfahrens;

Fig. 2 einen in dieser Vorrichtung vorgesehenen bildgebenden Spektrographen mit integrierter Aufnahmeeinheit;

Fig. 3 ein in dieser Vorrichtung vorgesehenes Förderband; und

Fig. 4 den Querschnitt dieses Förderbandes darstellen.

The invention is explained below by way of example with reference to FIGS. 1 to 4, wherein

1 shows a possible device for carrying out the method according to the invention;

2 shows an imaging spectrograph provided in this device with an integrated recording unit;

3 shows a conveyor belt provided in this device; and

Fig. 4 represent the cross section of this conveyor belt.

An advantageous embodiment of the method according to the invention for use as an acquisition, classification and control system for the sorted sorting of materials is subsequently based on the Exemplary representations according to FIGS. 1 to 4 explained in more detail:

The objects, objects and materials to be sorted are in Direction of arrow 7 on the conveyor belt 4 of the device 1 well distributed applied. After a short acceleration distance these items arrive at a speed of typically 1 to 3 m / s to the detection unit 2, which is above the Conveyor belt is arranged. The area of the conveyor belt to be considered is illuminated by at least one optical radiation source, which is preferably characterized by an over the observation area characterized spatially and spectrally homogeneous intensity distribution. In With this arrangement it is possible to create lines which run across the conveyor belt run through an optic 8 of the detection unit 2, as shown in 2 is to be detected. That from this line of observation outgoing light passes through the entry gap 9 - see Figure 2 - in an imaging spectrograph 10, with a wavelength-dispersive element 11, preferably a transmission grating, is equipped so that on the exit side of the spectrograph while maintaining the spatial information of the depicted Line the light rays of the lines shown spectrally broken down a recording unit 12 are mapped. This recording unit 12 is advantageously a two-dimensional array, such as a CCD (charge-coupled device) or CMOS (complementary metal-oxide semiconductor) Camera or any other camera which is suitable for the respective Spectral range is particularly suitable.

The data are transmitted to the evaluation unit 3 and there further processed. With the detection system according to the invention defines two categories of data at the same time, namely on the one hand the spectral data for determining the color and / or the chemical Properties and on the other hand the image data for determining shape, Location, size and / or structure of the objects to be classified. The shape of the objects is taken up by the lines across the conveyor belt with a typical resolution of between 128 Points and 1600 points are recorded and evaluated. Through the sequential recording of the lines arises from the aforementioned Lines a two-dimensional image. From this data set can be the geometric properties of the objects to be classified determine.

The data processing includes for typical, purely qualitative classification a dark current correction, a referencing against one Standard, a normalization of the spectra, a filter stage for Reduction of noise, the calculation of the 1st or 2nd derivative as well as the actual classification according to one on the respective Use case coordinated and optimized algorithm. Further For example, functions can be used to determine the center of gravity and / or the object or class edges and / or others Object properties by means of image processing algorithms. This makes it possible to precisely cut the objects to be separated recognize, classify in real time and further, for example make an appropriate sorting.

In the case of waste separation, for example, is for recycling Plastic bottles, such as polyethylene terephthalate (PET) bottles, not just the classification according to material, components, but also by color and / or by the type of screw cap of the bottle for the further work advantageous. Standard sorting criteria, that with other classification methods Find use, such as length, width and shape or contour directly the data are extracted and used for sorting. By Combination of the required data in the software evaluation module, for example, by comparing with deposited Target specifications, the classification result can be related to the process-specific Requirement can be adjusted. The data is the so-called Output data of the evaluation module are sent via a custom interface, another model of computer software, to the corresponding control components, for example a higher-level process control system or, as in this exemplary application, to a unit for automated sorting, such as controllable blow-out nozzles. At this Application for material sorting can thus be determined an automated sorting of the aforementioned data, for example by blowing out or by means of controlled flaps.

Sorting with the aid of a controllable one is particularly preferred Row of nozzles 5, in which air nozzles 14 are arranged such that the air flows generated thereby the objects to be sorted out Swirl diagonally to the side in the appropriate collection container so that these can be pulled off the conveyor belt 4. The nozzle row 5 is, as can be seen from Figures 3 and 4, advantageously below arranged the dreams of the strips 13 carrying the objects.

Another preferred method is to use an additional one Cross conveyor belt 6, which is at a distance of, for example, 30 cm is attached across the conveyor belt 4. The to be carried out Objects are in this embodiment by the Blow nozzles whirled onto this conveyor belt and then up transported away from the cross conveyor belt. This sorting process will explained in more detail based on the representation according to FIG. 3. This is too can see that the conveyor belt 4 in the longitudinal direction from several of each other spaced, e.g. there are about 50 strips 13. The row of nozzles 5 is mounted so that the individual nozzles 14 between the Strip 13 of the conveyor belt come into effect. At these blow-out stations the objects to be sorted are placed under the Conveyor belt mounted, controllable row of nozzles 5, whereby by the special design of the conveyor belt in the form of the strips 13 no interruption of the tape and therefore no interruption in the Flow occurs. The shape of the nozzles 14 and what is produced thereby Flow profiles are set so that the ones to be sorted out Objects according to their weight and the air flow generated on the cross conveyor belt 6 or in a side container (not shown) are whirled. By using the special conveyor belt 4, for example a belt belt, and that Attaching the row of nozzles 5 under this band, those objects remain which are not yet sorted out in this separation stage, continue with the same position and location on the conveyor belt for the Transport to the next separation stage. This allows depending on the requirement profile several separation steps can also be carried out. Thereby is the automated sorting of different objects enables so that a specific sorting by composition, like different plastics, metals and the like, but also according to environmental impact, sales value or Calorific value is enabled.

In summary it can be said that the method according to the invention for identifying, classifying and sorting objects is therefore feasible in a simple manner, because of one Unity simultaneously and with high spatial and spectral resolution real-time geometric and spectral data, for example regarding color and / or chemical properties of the considered Objects to be measured. The method according to the invention is particularly relevant for measurement objects, which result from process engineering or other reasons relative to the recognition system in Movement. The method according to the invention can in particular be using the detection system according to the invention from a unit for the simultaneous determination of image and Spectral data and an associated unit for evaluation perform the determined image and spectral data in a simple manner.

Claims (19)

Process for identifying, classifying and sorting objects, objects and materials, characterized in that in one process step the spectral data, such as color, chemical properties, and spatial data, such as shape, size, position and / or structure, moved relative to the measuring unit or moving objects, objects and materials in both flat dimensions with high spatial and full spectral resolution are determined and recorded in real time and that the objects, objects and materials recorded in this way are classified based on this data in real time and then sorted. A method according to claim 1, characterized in that the spectrally and two-dimensionally spatially high-resolution spectral information is determined in real time by illumination and / or detection in the UV and / or in the visible and / or in the near infrared and / or in the middle infrared. Method according to Claim 1 or 2, characterized in that the spatially resolved spectral information is determined with the aid of a combination of an imaging spectrograph (10) and a recording unit (12) which is connected to an evaluation unit (3). Method according to claim 3, characterized in that the evaluation of the determined image and spectral information takes place in real time with the aid of a computer-based classification program in the evaluation unit (3). Method according to one of claims 1 to 4, characterized in that the sorting and / or classification criteria are selected and / or set by means of a software module. Method according to one of claims 1 to 5, characterized in that the objects corresponding to a sorting criterion are withdrawn from the process cycle by means of a sorting device, preferably by means of a controllable air flow, without interrupting the overall process. Detection system for carrying out a method according to a of claims 1 to 6, consisting of a detection unit (2) for the simultaneous determination of spatially resolved Image and spectral data in real time, which is above the a transport system, preferably a conveyor track (4), distributed, classified and / or sorted objects is positioned, and one connected to this unit Evaluation unit (3) for evaluating the determined multispectral Image data. Detection system according to claim 7, characterized in that the object to be classified is stationary and that the detection unit (2) is mounted on a single or multi-axis movement system and is guided over the object. Detection system according to one of claims 7 or 8, characterized in that the detection unit (2) is an imaging spectrograph (10) in combination with a recording unit (12), whereby lines which are transverse to the direction of relative movement between the observed object, object or The material and the detection unit (2) run, imaged with the aid of suitable, spectrally tuned optics (8) and fed to the imaging spectrograph (10) via an entry slit (9), which is equipped with a wavelength-dispersive element, preferably a transmission grating (11), is equipped for extracting the spectral information from a light beam, so that on the exit side of the spectrograph while maintaining the spatial information of the depicted line, the light beams of the depicted lines are spectrally broken down and imaged on the recording unit (12). Detection system according to one of Claims 7 to 9, characterized in that the evaluation unit (3) is a data processing device for carrying out the various mathematical algorithms for preprocessing, classification and postprocessing of the multispectral image data in real time. Arrangement for executing the method according to one of claims 1 to 6, characterized by the transfer of the classification and image processing results of the evaluation unit (3) to a higher-level process control system, such as process control parameters, parameters for monitoring the process state, parameters for the detection of operational faults and / or faulty ones Objects and / or missing parts of objects. Arrangement for carrying out the method according to one of Claims 1 to 6, characterized by a continuous conveyor track (4) and at least one device (5) which is effectively connected to the detection system according to one of Claims 7 to 9 for sorting out those recognized by the detection system as corresponding to the sorting criteria Objects, objects or materials. Arrangement according to claim 12, characterized in that the sorting out is controlled by means of computer software on the basis of the data determined by the evaluation unit (3). Arrangement according to claim 12 or 13, characterized in that the device for sorting out comprises a row of nozzles (5) and laterally attached collecting containers. Arrangement according to claim 12 or 13, characterized in that the device for sorting out comprises a row of nozzles (5) and a transverse conveyor belt (6) positioned transversely over the conveyor track (4). Arrangement according to one of claims 12 to 15, characterized in that the conveyor track (4) is continuous over the entire system. Arrangement according to claim 16, characterized in that the continuous conveyor track (4) is formed by a plurality of parallel strips (13) in the longitudinal direction and that the row of nozzles (5) is arranged below the dreams of the strips (13) carrying the objects. Arrangement according to one of claims 12 to 17, characterized in that the objects to be sorted out are removed obliquely laterally from the material flow by means of blow-out. Arrangement according to claim 18, characterized in that the objects to be sorted out are whirled out onto a transverse conveyor belt (6) located above the conveyor belt and are thus removed from the material flow.

Images