CN102213591A - Digital image analysis device - Google Patents

Abstract

The present invention provides a digital image analysis device, which includes: a device for receiving a plurality of line settings related to an image edge of a digital image; and a device for defining a plurality of image positions where the plurality of lines intersect with the image edge of the digital image. When a computer executes the functional modules in the aforementioned digital image analysis device to perform digital image recognition, it only needs to analyze the pixel values on the path of each line, and does not need to analyze all pixels in the digital image, which greatly reduces the computing resources required by the computer's processor.

Description

Digital Image Analysis Device

technical field

The present invention relates to the technique of digitally analyzing images, especially a method related to digital image identification and measurement and a digital image analysis device.

Background technique

Digital image recognition technology has a wide range of applications. For example, in many applications such as product inspection, microscopic measurement, and image object recognition, image recognition technology can be used to judge the edge and shape of specific parts in digital images. image features.

However, the conventional digital image recognition method needs to use a computer processor to compare and analyze the pixel values of all the pixels of the digital image, which consumes a lot of computing resources. Therefore, the processing speed of the conventional digital image recognition method is not only limited by the computing power of the processor of the computer, but also difficult to execute on devices with low processor computing power.

In addition, conventional image recognition methods are also likely to affect the accuracy of image recognition due to the relationship between image content (such as image texture, image shape, etc.).

In view of this, how to improve or solve the deficiency of the above-mentioned digital image recognition method in related fields is a problem to be solved.

Contents of the invention

This specification provides a digital image analysis device, which includes: a device for receiving a plurality of line settings related to an image edge of a digital image; and a device for defining the plurality of lines and the image edge of the digital image A device for intersecting multiple image positions.

One of the advantages of the aforementioned digital image analysis device is that when the processor of the computer performs digital image recognition, it only needs to analyze the pixel values on the path of each line, instead of analyzing all the pixels in the digital image, which greatly reduces the processing time. Computing resources required by the processor.

Description of drawings

FIG. 1 is a simplified functional block diagram of an embodiment of the digital image analysis device of the present invention.

FIG. 2 is a flowchart of an embodiment of the digital image recognition method of the present invention.

FIG. 3 and FIG. 4 are examples of display screens of the display in FIG. 1 .

FIG. 5 is a partially enlarged view of the digital image in FIG. 4 .

6 and 7 are examples of display screens of the display in FIG. 1 .

FIG. 8 is a flowchart of an embodiment of the digital image measurement method of the present invention.

FIG. 9 is an embodiment of a display screen of the display in FIG. 1 .

[Description of main component symbols]

100  Digital image analysis device 102 subject matter 110 Image capture device 120 Host 122 Processor 124 storage module 130 display 140 input interface 150 Carrying device 300, 600, 700, 900 digital imaging 302, 602 opening 304, 306, 604, 606 side 310 option column 410, 420, 430, 440, 450, 460, 470, 702, 704, 706, 720, 730, 740, 750, 760, 770, 780, 790, 910, 920, 930, 940, 950, 960 lines 412, 422, 432, 442, 452, 462, 472, 552, 554, 556, 612, 622, 632, 642, 652, 662, 672, 712, 714, 716, 722, 732, 742, 752, 762, 772, 782, 792, 912, 922, 932, 942, 952, 962 Image location 502, 504 line endpoint 510, 512, 514, 516, 518, 520, 522, 524, 526, 528, 530, 532, 534, 536, 538, 540 pixel 710 arc

Detailed ways

Embodiments of the present invention will be described below in conjunction with related figures. In the drawings, the same reference numerals designate the same or similar elements.

Certain terms are used in the specification and following claims to refer to particular elements. Those skilled in the art should understand that the same element may be called by different terms. This description and the subsequent claims do not use the difference in name as the way to distinguish components, but the difference in function of the components as the basis for distinction. "Includes" mentioned throughout the specification and subsequent claims is an open term, so it should be interpreted as "including but not limited to...".

FIG. 1 is a simplified functional block diagram of a digital image analysis device 100 according to an embodiment of the present invention. As shown in FIG. 1 , the digital image analysis device 100 includes an image capture device 110 , a host 120 , a display 130 , and an input interface 140 . The image capturing device 110 , the display 130 and the input interface 140 are coupled to the host 120 . The term "coupled" in this specification and subsequent claims includes any direct and indirect means of connection. Therefore, the image capture device 110, the display 130, and the input interface 140 can be directly connected to the host 120 (including through electrical connection or signal connection methods such as wireless transmission and optical transmission), or indirectly through other devices or connection means. It is electrically or signally connected to the host 120 .

The image capturing device 110 is used to sense the image of the target object 102 placed on the carrying device 150 . In practice, the image capturing device 110 may be a digital camera, a digital video camera, a digital microscope or any other device capable of image sensing. In this embodiment, the host 120 includes a processor 122 and a storage module 124 . The storage module 124 can be realized by one or more storage media.

The digital image analysis device 100 can be used for digital image recognition or digital image measurement. The operation of the digital image analysis device 100 will be further described below with reference to FIG. 2 .

FIG. 2 is a flowchart 200 of an embodiment of the digital image recognition method of the present invention.

In step 210 , the processor 122 of the host 120 displays the digital image of the target object 102 sensed by the image capture device 110 on the display 130 . For example, the digital image 300 shown in FIG. 3 is a partial image of a circuit board sensed by the image capture device 110, wherein, reference numeral 302 represents an opening on the circuit board, and reference numerals 304 and 306 represent the circuit board. Both sides of the circuit board.

For digital image recognition processing, the recognition of image edges is a very important part. After defining the position of the edge of the image, the shape of the image can be further identified. In order to reduce the computing resources required for digital image recognition processing and improve the speed and accuracy of recognition, the digital image analysis device 100 will interact with the user during the process of digital image recognition, requiring the user to provide information that can help determine the edge of the image and /or auxiliary information for the shape. The following will take the image recognition process of the opening 302 of the circuit board as an example to illustrate.

In step 220, the processor 122 displays a plurality of options for identifying patterns in the form of text or patterns on the display 130 for the user to choose. For example, as shown in FIG. 3 , the processor 122 can use a plurality of options for identifying patterns (such as circles, ellipses, polygons, irregular shapes, arcs, straight lines, polylines, etc.) The form groups form option bar 310 and are displayed on display 130 . Each recognition pattern option has a corresponding image edge calculation method, which can be stored in the storage module 124 in the form of a computer program. The content and number of the aforementioned identification pattern options are only an example, and are not intended to limit the actual implementation of the present invention. In practice, the options and content of the identification pattern can be increased, decreased or changed according to the needs of the design.

In step 230 , the user needs to select one or more recognition patterns related to the shape of the digital image 300 from a plurality of recognition pattern options presented in the option bar 310 . In practice, the user can select one or more identification patterns through the input interface 140 by means of cursor selection, touch, voice control, or other command input methods. In this embodiment, the user can select the recognition pattern "circle" that is closest to the shape of the circuit board opening 302 .

In step 240 , the host 120 receives the recognition pattern setting selected by the user through the input interface 140 and records it in the storage module 124 as a reference for the processor 122 to perform image recognition.

In step 250 , the user needs to roughly set a plurality of simple lines on the digital image 300 . The user can set a plurality of lines corresponding to the image edge of the opening 302 to be identified in the digital image 300 by clicking or touching the input interface 140 . The number of lines required is related to the identification pattern selected by the user. For example, when the selected identification pattern is "circle" or "arc", at least 3 lines are required. When the selected identification pattern is "Straight Line", at least 2 lines are required. When the selected identification pattern is "polygon" or "irregular shape", a larger number of lines is required.

Each line only needs to intersect with the image edge of the image object to be identified, that is, the two ends of each line should be respectively located on both sides of the image edge. The length of each line is not limited, and the shape is not limited to a straight line, and may be a curved line or an irregular line. In order to reduce the complexity of subsequent calculations, each line preferably only intersects with a single image edge. In practice, the display 130 and the input interface 140 can be integrated together, for example, a touch screen can be used to realize the functions of the display 130 and the input interface 140 .

As shown in the embodiment of FIG. 4 , the user can roughly draw three lines 410 , 420 , and 430 on the digital image 300 that intersect the image edge of the opening 302 .

In step 260 , the host 120 receives the setting data of multiple lines input by the user through the input interface 140 , and records them in the storage module 124 as a reference when the processor 122 performs image edge recognition.

In step 270, the processor 122 reads the setting data of a plurality of lines recorded in the storage module 124, and defines a plurality of images where the plurality of lines intersect with the image edge of the opening 302 in the digital image 300 Location. For example, the processor 122 in this embodiment will calculate the image position 412 where the line 410 intersects the image edge of the opening 302 , the image position 422 where the line 420 intersects the image edge of the opening 302 , and the distance between the line 430 and the opening 302 Image location 432 where image edges intersect.

There are many methods for the processor 122 to calculate the image position where each line intersects the edge of the image of the aperture 302 . In one embodiment, the processor 122 may calculate the pixel value difference between adjacent pixels on the path corresponding to each line, and set the position where the pixel value difference is the largest as the intersection position between the line and the image edge of the opening 302 . This calculation of image position can be used, for example, in applications where there are significant differences in pixel values on either side of the edge of the image.

Please note that the pixel value referred to in the description and subsequent claims can be implemented as the luminance value (luminance) or chroma value (chrominance) of the pixel.

In another embodiment, the processor 122 may sequentially calculate the pixel value difference between adjacent pixels along a predetermined direction on the path corresponding to each line, and calculate the pixel value difference that reaches or exceeds a predetermined critical value. The first location is set as the intersection of the line with the edge of the image of the opening 302 . For example, in applications where there is no significant difference in pixel values on both sides of the edge of the image, but the pixel values on both sides of the edge of the image are significantly different from those on both sides of the edge of the image, this method of calculating the image position can be used. For convenience, the partial enlarged view of the digital image 300 shown in FIG. 5 is taken as an example for illustration below.

In FIG. 5 , pixels 510 , 512 , 514 , 516 , 518 , 520 , 524 , 526 , and 528 are local pixels on the image edge of aperture 302 . Pixels 530 , 532 , 518 , 516 , 534 , 536 , 538 , and 540 are pixels located on the path of line 410 . Numbers 502 and 504 represent the two endpoints of the line 410 . The user can fine-tune the intersection position of the line 410 and the edge of the image of the opening 302 by adjusting the position of one end point of the line 410 (for example, the end point 504 ) left or right. This way of setting the intersection position between a line and an image edge can provide users with higher options than using a mouse (or touch means) to directly select a specific point on the edge of an image. precision.

The processor 122 may follow the direction D1 to sequentially calculate the pixel value difference between two adjacent pixels on the path of the line 410 (for example, first calculate the pixel value difference between the pixels 530 and 532, then calculate the pixel value difference between the pixels 532 and 518, The difference between the pixel values of pixels 518 and 516 is then calculated, and so on), and compared with the predetermined threshold value. In this embodiment, the pixel value difference between pixels 532 and 518 is the first pixel value difference exceeding the predetermined critical value, therefore, the processor 122 may omit calculating subsequent pixel value differences, and combine the pixels 532 and 518 The intersection location 552 is set as the image location 412 where the line 410 intersects the image edge of the aperture 302 .

In another embodiment, the processor 122 may first calculate the median or the average value of the plurality of pixel values along the path of the line 410, and set the median or the average value as a critical value. In other words, the setting of the threshold can be dynamically adjusted. Next, the processor 122 can sequentially compare the pixel values on the path of the line 410 with the critical value along the direction D1, and set the first position where the pixel value reaches or crosses the critical value as the line 410 and the opening. 302 is the intersection position of the image edge. Assuming that the pixel 518 is the first pixel whose value reaches the critical value along the direction of D1 along the path of the line 410, the processor 122 may omit comparing the subsequent pixel values, and use the center position 554 of the pixel 518 as the line 410 and the open pixel. Image location 412 where image edges of holes 302 intersect.

Assuming that the pixel 518 is the first pixel whose pixel value crosses the critical value (that is, the pixel value is greater than or smaller than the critical value for the first time) along the path of the line 410 along the direction of D1, the processor 122 may omit comparing subsequent pixel value, and set the intersection position 552 of the pixel 518 and the pixel 532 as the image position 412 where the line 410 intersects the image edge of the opening 302 . Alternatively, the processor 122 can also interpolate the pixel value 512 and the pixel value of the pixel 532 to calculate the position 556 corresponding to the critical value, and use the position 556 as the image where the line 410 intersects with the image edge of the opening 302 Location 412.

It can be seen from the above description that the precision (Precision) when the processor 122 defines the intersection position of each line and the image edge of the image object can be smaller than one pixel unit.

Next, in step 280, the processor 122 connects the plurality of image positions obtained in step 270 according to the identification pattern selected in step 230, so as to identify one or more local edges or edges of the digital image 300 full border. For example, in this embodiment, the recognition pattern selected by the user in step 230 is "circle", so the processor 122 will determine the three image positions 412, 422 and 432 defined in step 270. A circle is determined as the image edge recognition result of the opening 302 .

After identifying the boundary of an image object (such as the circumference of the opening 302 ), the processor 122 can further calculate the location of the geometric feature of the image object (such as the center 408 of the opening 302 ) as needed.

Next, the digital image recognition method of the present invention will be described by taking the image recognition process of the circuit board side 304 as an example.

Since the circuit board side 304 is a linear side, the user can select the identification pattern "straight line" that is closest to the shape of the circuit board side 304 in step 230, and set it in step 250 to match the shape of the circuit board. Two lines 440 and 450 where the image edges of side 304 intersect.

In step 270, the processor 122 can use one of the aforementioned methods to define the image position 442 where the line 440 intersects the image edge of the circuit board side 304 and the image where the line 450 intersects the image edge of the circuit board side 304 Location 452.

In step 280, since the identification pattern recorded in the storage module 124 is "straight line", the processor 122 determines a straight line according to the image positions 442 and 452 as the image edge identification result of the circuit board side 304 .

In another embodiment, in step 230, the user can also select the recognition pattern "polyline" that is closest to the shape of both the side 304 and the side 306 of the circuit board, and in step 250 set Two lines 440 and 450 intersect the image edge of side 304 and two lines 460 and 470 intersect the image edge of circuit board side 306 . Afterwards, in step 270, the processor 122 defines the image position 442 where the line 440 intersects the image edge of the circuit board side 304, the image position 452 where the line 450 intersects the image edge of the circuit board side 304, the line 460 and the circuit board. The image location 462 where the image edge of the board side 306 intersects and the image location 472 where the line 470 intersects the image edge of the circuit board side 306 .

In this way, in step 280, the processor 122 can connect the image positions 442, 452, 462 and 472 to form an L-shaped fold line according to the indication of the recognized pattern "polyline", which is used as the edge of both sides 304 and 306 of the circuit board. Image recognition results.

After obtaining the position of the center 408 of the opening 302 and the positions of the sides 304 and 306 of the circuit board, the processor 122 can further calculate the distance from the center 408 to the side 304 or 306 of the circuit board.

As can be seen from the foregoing description, the digital image recognition method of the present invention only needs to analyze the pixel values on the path of each line to calculate the intersection position of the line and a specific image edge, thereby achieving the purpose of identifying the edge of the image without having to All the pixels in the digital image 300 are analyzed, which greatly reduces the computing resources required by the processor 122 . In addition, since the processor 122 uses the recognition pattern setting input by the user as a reference for image recognition, the accuracy of recognizing the edge or shape of the digital image can be greatly improved.

In terms of application, the digital image analysis device 100 can be used to achieve the purpose of product inspection and quality control. For example, the carrying device 150 can be realized by a conveyor belt on a production line, and is used to periodically deliver the target object 102 to be inspected (such as the aforementioned circuit board) to the inspection area where the image capture device 110 is aligned. There are often many similar image features between the target objects to be detected, so the user only needs to input appropriate auxiliary information (such as image recognition pattern, Lines related to the edge of the image object, etc.) to the host 120, the image capture device 110 can automatically complete the image recognition processing of the subsequent target object 102 according to the same auxiliary information, and then realize the screening operation of good or bad products. The example in FIG. 6 will be used for further description below.

In FIG. 6 , a digital image 600 is a partial image of another circuit board sensed by the image capture device 110, wherein, reference numeral 602 represents an opening on the circuit board, and reference numerals 604 and 606 represent two holes of the circuit board. side. Generally speaking, the differences between products under the same process are usually within a limited range. Therefore, the digital image analysis device 100 can retain the auxiliary lines 410, 420, 430, 440, 450, 460, and 470 input by the user in the process of recognizing the digital image 300 of the previous circuit board, and recognize the pattern "circle". and "polyline" (or "line") settings. The processor 122 can use one of the aforementioned digital image recognition methods to automatically define the image position 612 where the line 410 intersects the image edge of the opening 602 , the image position 622 where the line 420 intersects the image edge of the opening 602 , and the line 430 The image location 632 that intersects the image edge of the aperture 602 . Then, the processor 122 can determine the circumference of the opening 602 by using the three positions of the image positions 612 , 622 and 632 according to the setting of the “circle” recognition pattern, and then calculate the position of the center 608 of the opening 602 .

Similarly, the processor 122 can use one of the aforementioned digital image recognition methods to define the image position 642 where the line 440 intersects the image edge of the circuit board side 604 and the image where the line 450 intersects the image edge of the circuit board side 604 Location 652 , image location 662 where line 460 intersects the image edge of circuit board side 606 , and image location 672 where line 470 intersects the image edge of circuit board side 606 . Next, the processor 122 can connect the image positions 642 , 652 , 662 and 672 into an L-shaped broken line according to the setting of the recognition pattern “polyline” in step 280 as the image recognition result of the circuit board sides 604 and 606 .

Next, the processor 122 can further calculate the distance from the center 608 of the opening 602 to the side 604 and/or side 606 of the circuit board to determine whether the position of the opening 602 meets the preset specifications, and then determine the circuit board. Whether it is a qualified circuit board.

As can be seen from the above description, when the digital image recognition method of the present invention is applied to product inspection, it only needs the simple intervention of the user (for example, inputting the image recognition pattern and a few simple lines, etc.), and it can be almost fully automatic. To complete the testing work of subsequent products. Furthermore, since the aforementioned image recognition and product inspection processes require far less computing resources than conventional methods, the efficiency of the quality control system can be effectively improved and the required hardware costs can be reduced.

In addition to being used for product inspection, the digital image analysis device 100 can also apply the aforementioned digital image recognition method to the application of object positioning and alignment. For example, after the processor 122 of the host computer 120 recognizes the relative position and distance between the target object 102 and a specific position on the carrying device 150 (usually a design such as a positioning line, a positioning edge, a positioning angle, or a positioning point), the The carrying device 150 can be controlled to move and/or rotate the target object 102 to a predetermined position, so as to achieve the purpose of object positioning or alignment.

For a concave digital image, the aforementioned digital image recognition method can still successfully recognize the local edge or the entire boundary of the digital image. For example, for the digital image 700 shown in FIG. A plurality of lines 720 , 730 , 740 , 750 , 760 , 770 , 780 , and 790 intersect the edges of the image of the image. In this way, the processor 122 can perform step 270 according to these settings to calculate the image positions 722 , 732 , 742 , 752 , 762 , 772 , 782 and 792 where the lines intersect with the image edges of the digital image 700 .

In addition, the user can perform the aforementioned digital image recognition method again, but instead select the recognition pattern “arc” and set the lines 702 , 704 , and 706 intersecting the image edge of the arc 710 . In this way, the processor 122 can identify the image locations 712 , 714 , and 716 where the lines 702 , 704 , and 706 intersect the image edges of the arc 710 .

After obtaining the aforementioned image positions 712, 714, 716, 722, 732, 742, 752, 762, 772, 782, and 792, the processor 122 can further identify the pattern "polygon" and "arc" according to the aforementioned settings. Then proceed to step 280, determine a plurality of image edge positions of the digital image 700 according to these image positions, and then quickly recognize the complete shape of the digital image 700.

For the digital image of irregular shape, the user can choose to identify the pattern as "irregular shape" and set more lines intersecting with the edge of the digital image of the irregular shape. The more lines the user sets, the more accurate the processor 122 can improve the shape recognition of the digital image.

Please note that the execution sequence of the steps in the aforementioned flow chart 200 is only an example, rather than limiting the implementation of the present invention. For example, the order of step 230 and step 250 can be reversed. In addition, the digital image analysis device 100 can also be designed to be specially used for recognizing digital images of objects of the same shape, and in this case, steps 220 , 230 and 240 can be omitted.

The digital image analysis device 100 can further apply the aforementioned digital image recognition method to digital image measurement. For example, FIG. 8 shows a flowchart 800 of an embodiment of the digital image measurement method of the present invention. Steps 210 to 280 in the flowchart 800 are the same as the processes with the same numbers in the aforementioned flowchart 200 , and for the sake of brevity, details are not repeated here.

As shown in FIG. 8 , after the processor 122 defines a plurality of image positions on the digital image in step 270, it can further execute step 870, according to the scaling scale of the digital image, the distance between two specific image positions on the digital image, or more The total length of the path of image positions, converted from the pixel distance to the actual length value.

In addition, after the processor 122 recognizes the partial edge or the entire boundary of the digital image in step 280, it may also proceed to step 880 to measure specific image feature values of the digital image according to the image edge recognition result of the digital image. The following uses FIG. 9 as an example for further description.

The digital image 900 in FIG. 9 is a cross-sectional image of a circular tube. If the target to be measured is an image characteristic value related to the image inner edge of the digital image 900, such as the inner diameter size, inner ring cross-sectional area, etc., then the user can select a shape close to the image inner edge shape of the digital image 900 in step 230. The pattern "circle" is recognized, and in step 250, a plurality of lines intersecting the inner edge of the digital image 900 are set according to a certain direction. For example, in the embodiment of FIG. 9, the lines 910, 920, and 930 are set from the inside out.

Afterwards, in step 270, the processor 122 only needs to analyze the pixel values along the drawing direction of each line on the path of each line according to the above-mentioned calculation method, so as to define where the line 910 intersects with the image inner edge of the digital image 900 Image location 912 , image location 922 where line 920 intersects the inner edge of the image, and image location 932 where line 930 intersects the inner edge of the image. That is, the aforementioned pixel value analysis operation is performed along the direction D1 on the path of the line 910 , along the direction D2 on the path of the line 920 , and along the direction D3 on the path of the line 930 .

In step 280 , the processor 122 can use the image positions 912 , 922 and 932 to determine a circumference corresponding to the recognition pattern “circle” as the recognition result of the image inner edge of the digital image 900 .

After identifying the circumference of the inner edge of the digital image 900, the processor 122 can further calculate image feature values related to the inner edge of the digital image 900 according to the scaling scale of the digital image 900 in step 880, for example, the circular tube The inner diameter of the section, the radius, circumference, cross-sectional area, average color, etc. of the inner ring of the tube.

Similarly, if the object to be measured is the image feature value related to the image outer edge of the digital image 900, such as the outer diameter size, the cross-sectional area of the outer ring of the circular tube, etc., the user only needs to set it according to a certain direction in step 250. A plurality of lines intersecting the image outer edge of the digital image 900 is defined. For example, in the embodiment of FIG. 9, the lines 950, 960 and 970 are set from outside to inside. In step 270, the processor 122 will perform pixel value analysis along the direction D5 on the path of the line 950, perform pixel value analysis along the direction D6 on the path of the line 960, and perform pixel value analysis along the direction D7 on the path of the line 970. Pixel value analysis is performed to define the image location 952 where the line 950 intersects the outer edge of the image, the image location 962 where the line 960 intersects the outer edge of the image, and the image location 972 where the line 970 intersects the outer edge of the image.

Then, in step 280 , the processor 122 can use the image positions 952 , 962 and 972 to identify the circumference of the image outer edge of the digital image 900 .

Therefore, in step 880, the processor 122 can further calculate image feature values related to the image outer edge of the digital image 900 according to the scaling scale of the digital image 900, such as the outer diameter of the circular tube section, the outer diameter of the circular tube Data on circle radius, circumference, cross-sectional area, average color, etc.

Using the aforementioned digital image measurement method, the user only needs to provide a small amount of auxiliary information, and the digital image analysis device 100 can quickly and accurately measure the specific geometrical feature (Geometrical feature) value of the image object, for example, the length of the local edge , arc length, radian, included angle, central angle, circumference angle, chord angle, perimeter, area, radius, diameter, inner diameter, outer diameter, etc. of the image object, or image features such as the average color and average brightness of the image object value.

In practice, in addition to identifying the shape of a specific image part in the digital image, the processor 122 can also be equipped with an appropriate image recognition mechanism to further identify and calculate the specific image features (such as particles, crystal structures, etc.) contained in the image part. location and/or quantity, allowing users to perform more applications.

As mentioned above, the precision of the processor 122 in defining the intersection position of each line and the image edge of the image object can be smaller than a pixel unit. Therefore, when the aforementioned digital image recognition method is applied to digital image measurement, the accuracy of image measurement can be effectively improved.

Please note that each component in the device claim in the claims corresponds to each step of the aforementioned computer program flow. Therefore, the device claim in the claims should be understood as realizing the aforementioned computer program mainly through the computer program described in the specification. The functional module architecture of the solution.

The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the claims of the present invention shall fall within the scope of the present invention.

Claims (15)

1. digital image analysis device, it includes:

Be used to receive the device that a plurality of lines relevant with the image edge of a digitized video are set; And

Be used to define the device of a plurality of image positions that the image edge of these a plurality of lines and this digitized video intersects.

2. digital image analysis device as claimed in claim 1, it includes in addition:

Be used to connect the device of these a plurality of image positions, with at least one local edge of this digitized video of identification.

3. digital image analysis device as claimed in claim 2, it includes in addition:

Be used to show the device of the option of a plurality of identification type samples; And

Be used to receive the device that selected one or more identification type sample of a user is set.

4. digital image analysis device as claimed in claim 3, the device that wherein is used to connect these a plurality of image positions includes:

Be used for setting the device that connects these a plurality of image positions according to this one or more identification type sample.

5. digital image analysis device as claimed in claim 3, the device that wherein is used to connect these a plurality of image positions includes:

Be used for determining the device of plural edges line according to this one or more identification type sample setting and these a plurality of image positions; And

Be used for determining one or more local edge of this digitized video or the device of integral edge according to this plural edges line.

6. digital image analysis device as claimed in claim 3, wherein this one or more identification type sample is set relevant with the shape of this digitized video.

7. digital image analysis device as claimed in claim 3, wherein one or more identification type sample of the quantity and this of these a plurality of lines is set corresponding.

8. digital image analysis device as claimed in claim 2, it includes in addition:

Be used at least one local edge, judge the device of position of a geometric properties of this digitized video according to this digitized video.

9. digital image analysis device as claimed in claim 2, it includes in addition:

Be used for identification result, calculate the geometrical characteristic of this digitized video or the device of image feature value according to image edge.

10. digital image analysis device as claimed in claim 1, it includes in addition:

Be used for calculating the device of the spacing or the path of these a plurality of image positions according to the scaling chi of this digitized video.

11. digital image analysis device as claimed in claim 1, the device that wherein is used to define these a plurality of image positions includes:

Be used to analyze the pixel value of the pixel on each lines respective path, with the device of the intersection location of the image edge that determines these lines and this digitized video.

12. digital image analysis device as claimed in claim 11, the device that wherein is used to analyze the pixel value of the pixel on each lines respective path includes:

Be used to calculate the device of the value differences between neighbor on each lines respective path; And

Be used for the set positions of value differences maximum device for the intersection location of the image edge of these lines and this digitized video.

13. digital image analysis device as claimed in claim 11, the device that wherein is used to analyze the pixel value of the pixel on each lines respective path includes:

Be used for calculating the device of the value differences between neighbor in regular turn along a predetermined direction in the path of each lines correspondence; And

Be used for value differences is met or exceeded the device of first set positions of a predetermined critical for the intersection location of the image edge of these lines and this digitized video.

14. digital image analysis device as claimed in claim 11, the device that wherein is used to analyze the pixel value of the pixel on each lines respective path includes:

Be used in the path of each lines correspondence the device that in regular turn pixel value on this path and a critical value is compared along a predetermined direction; And

First set positions that is used for pixel value reached or crosses over this critical value is the device of intersection location of the image edge of lines and this digitized video.

15. digital image analysis device as claimed in claim 14, the device that wherein is used to analyze the pixel value of the pixel on each lines respective path includes:

Be used for setting the device of the pairing critical value of these lines according to the pixel value on the path of each lines correspondence.

Images