JP4990960B2 - Object identification device, object identification method, and object identification program - Google Patents

Description

  The present invention relates to an object identification device, an object identification method, and an object identification program.

  Algorithms such as SIFT (Scale-invariant feature transform) and SURF (Speeded Up Robust Features) that extract and recognize feature points and feature quantities have attracted attention as prescriptions for identifying objects from images and videos. These methods have the feature that the recognition rate of an object is not easily lowered even when the object is shielded, the object rotates, the object scale changes, the brightness of the illumination lamp changes with respect to the object (for example, Non-patent document 1).

David G. Low, "Object Recognition from Local Scale-Invariant Features", Proceedings of the International Conference on Computer Vision, 1999. 9, pp. 1150-1157

  However, in the conventional technique, there is a problem that the recognition rate of an object is reduced when it is difficult to extract many feature points because the image has a small luminance difference and is flat. Further, since the amount of calculation for extracting feature points is very large, there is a problem that it is difficult to mount on a mobile terminal device or the like that cannot perform many calculations.

  The present invention has been made in view of the above problems, and provides an object identification device, an object identification method, and an object identification program capable of obtaining high recognition accuracy even in an image in which feature points are difficult to extract. The challenge is to do.

  To achieve the above object, according to the present invention, in the object identification device, a background difference processing unit that extracts a figure included in the input image from the input image by a background difference method, and a figure extracted by the background difference processing unit. A binarization unit that binarizes image data, a contour extraction unit that extracts contour data from graphic image data binarized by the binarization unit, and a centroid position of the contour data extracted by the contour extraction unit Calculating the distance from the center of gravity calculated by the center of gravity calculating unit to the outermost periphery of the contour data extracted by the contour extracting unit for each predetermined angle, and calculating the contour data as polar coordinates Polar coordinate values stored in the polar coordinate value storage unit, comprising a polar coordinate value generation unit that converts a representation into polar coordinate values and a polar coordinate value storage unit that stores polar coordinate values of contour data of a figure to be identified And said A polar representation value coordinate value generating unit has generated compared while shifting by a predetermined angle, it is characterized by performing the identification of the figure included in the input image.

  In the object identification device, the present invention also extracts an inner contour for extracting a black pixel region having a predetermined area or more existing inside the contour data extracted by the contour extraction unit and extracting inner contour data of the extracted region. The polar coordinate value generation unit calculates a distance from the centroid position calculated by the centroid calculation unit to the outermost peripheral portion of the inner contour data extracted by the inner contour extraction unit for each predetermined angle. Then, the inner contour data is converted into a polar coordinate representation to generate polar coordinate values, and the polar coordinate values of the inner contour data stored in the polar coordinate value storage unit and the inner contour data generated by the polar coordinate value generation unit The polar coordinate expression value is compared with each other while shifting by a predetermined angle, and the figure included in the input image is identified.

  In the object identification device, the present invention further includes a color space separation unit that separates image data of a graphic extracted from the input image into a plurality of color space components, and the binarization unit includes the color space. Each color space component separated by the separation unit is binarized, and the contour extraction unit extracts contour data from the image data of each color space component binarized by the binarization unit, and is stored in the polar coordinate value storage unit The polar coordinate value of each contour data of each color space component is compared with the polar coordinate expression value of each contour data of each color space component generated by the polar coordinate value generation unit while shifting by a predetermined angle and included in the input image It is characterized in that the figure to be identified is identified.

  In the object identification device according to the present invention, the polar coordinate value generation unit extracts a maximum distance value from the generated polar coordinate values, and performs normalization based on the extracted maximum distance value. It is said.

  According to the present invention, in the object identification method of the object identification device, the background difference processing unit extracts a figure included in the input image from the input image by the background difference method, and the binarization unit includes: A binarization step for binarizing the graphic image data extracted by the background difference processing step, and a contour extracting unit for extracting contour data from the graphic image data binarized by the binarization step An extraction step, a centroid calculation step in which the centroid calculation unit calculates a centroid position of the contour data extracted in the contour extraction step, and a polar coordinate value generation unit calculates the contour extraction step from the centroid position calculated in the centroid calculation step. A polar coordinate value generating step of calculating a distance to the outermost peripheral portion of the extracted contour data for each predetermined angle, converting the contour data into a polar coordinate expression to generate a polar coordinate value, and a contour of a figure to be identified Data A polar coordinate value storage unit that stores coordinate values, and compares the polar coordinate value stored in the polar coordinate value storage unit with the polar coordinate expression value generated by the polar coordinate value generation unit while shifting by a predetermined angle, It is characterized by identifying a figure included in the input image.

  Further, according to the present invention, in the object identification program of the object identification device, a background difference processing step for extracting a figure included in the input image from the input image by a background difference method and a background difference processing step are extracted from the input image. A binarization step for binarizing graphic image data, a contour extraction step for extracting contour data from the graphic image data binarized by the binarization step, and the contour data extracted by the contour extraction step A center-of-gravity calculation step for calculating a center-of-gravity position; and a distance from the center-of-gravity position calculated by the center-of-gravity calculation step to the outermost peripheral portion of the contour data extracted by the contour extraction step for each predetermined angle; A polar coordinate value generation step for generating a polar coordinate value by converting into a polar coordinate representation, and a polar coordinate value storage unit for storing a polar coordinate value of contour data of a figure to be identified, the polar coordinate The polar coordinate value stored in the storage unit and the polar coordinate expression value generated by the polar coordinate value generation unit are compared while being shifted by a predetermined angle, and the figure included in the input image is identified. Yes.

  According to the present invention, even when a large number of feature points cannot be extracted from an object, an appropriate contour is extracted from an input image using background difference processing, the extracted contour data is expressed in polar coordinates, and an object expressed in polar coordinates is expressed. High recognition accuracy can be obtained by performing recognition.

It is a block diagram showing an example of composition of an object discernment device concerning a first embodiment. It is a figure explaining a background difference process. It is a figure explaining the process of the input image data which concerns on the embodiment. It is a figure explaining an example of the generation method of the polar coordinate value concerning the embodiment. It is an example of the structure of the polar coordinate value which the polar coordinate value memory | storage part which concerns on the embodiment memorize | stores. It is a flowchart of the registration procedure of the target object concerning the embodiment. It is a flowchart of a polar coordinate value generation procedure according to the embodiment. It is a flowchart of the identification procedure of the target object concerning the embodiment. It is a figure explaining the identification method concerning the embodiment. It is a block diagram which shows an example of a structure of the outline extraction apparatus which concerns on 2nd embodiment. It is a figure explaining the color space channel decomposition | disassembly of the RGB image data which concerns on 2nd embodiment. It is an example of the structure of the polar coordinate value which the polar coordinate value memory | storage part which concerns on 2nd embodiment memorize | stores. It is a flowchart of the registration procedure of the target object which concerns on 2nd embodiment. It is a flowchart of the identification procedure of the target object which concerns on 2nd embodiment. It is a figure which shows the usage example of the object identification apparatus by this embodiment.

  Hereinafter, embodiments of the present invention will be described in detail with reference to FIGS. In addition, this invention is not limited to the embodiment which concerns, A various change is possible within the range of the technical thought.

[First embodiment]
The configuration of the object identification device according to the present embodiment will be described with reference to a block diagram illustrating an example of the configuration of the object identification device in FIG. The object identification device 1 includes an external setting value capturing unit 11, an image data capturing unit 12, a background image storage unit 13, a background difference processing unit 14, a smoothing unit 15, a binarizing unit 16, and an outline extraction. Unit 17, center-of-gravity calculation unit 18, inner contour extraction unit 19, polar coordinate value generation unit 20, identification unit 21, polar coordinate value storage unit 22, and identification result output unit 23. In addition, a camera 2 and an image display device 3 are connected to the object identification device 1.

  The camera 2 is composed of, for example, a light receiving lens and a CCD camera, captures a background image or input image that does not include an object, and transmits the captured input image to the object identification device 1. The image display device 3 receives the image data for displaying the contour extraction result from the object identification device 1 and displays the received image data.

  The external setting value capturing unit 11 captures external setting values that are various settings used for object identification processing from the outside. The external set value includes various parameters of smoothing performed by the smoothing unit 15 (processing method, number of processes, range to be processed, shape to be processed, weighting coefficient, etc.) and the like. In addition, the external setting value capturing unit 11 outputs the captured external setting value to the background difference processing unit 14, the smoothing unit 15, the binarization unit 16, and the polar coordinate value generation unit 20.

  The image data capturing unit 12 captures a background image or an input image captured by the camera 2, and writes and stores image data of the captured background image (hereinafter referred to as background image data) in the background image storage unit 13. Further, the image data capturing unit 12 outputs image data of the captured input image (hereinafter referred to as input image data) to the background difference processing unit 14.

  The background image storage unit 13 stores background image data. The background image data may be stored in the background image storage unit 13 in advance.

  The background difference processing unit 14 captures the external setting value output from the external setting value capturing unit 11 and the input image data output from the image data capturing unit 12, and the background image data stored in the background image storage unit 13. Is read. Further, the background difference processing unit 14 reads out various parameters related to the background difference processing from the fetched external setting values. Further, the background difference processing unit 14 calculates difference image data between the input image data and the background image data using various parameters related to the read background difference processing, and outputs the calculated difference image data to the smoothing unit 15. FIG. 2 is a diagram for explaining the background difference process. As shown in FIG. 2, the difference image data is obtained by calculating the difference between the background image data (FIG. 2 (a)) that does not include the object photographed in advance and the input image data (FIG. 2 (b)) that includes the object. (FIG. 2 (c)) is extracted.

  The smoothing unit 15 captures the external setting value output from the external setting value capturing unit 11 and the difference image data output from the background difference output unit 14. Further, the smoothing unit 15 reads the type of smoothing method and various parameters for smoothing from the acquired external setting values. Further, the smoothing unit 15 smoothes the acquired difference image data by the number of times of the acquired external setting value using, for example, a Gaussian filter based on the type of the read smoothing method. Further, the smoothing unit 15 performs a hole filling process on a region having a predetermined area or less using a general method. For example, the smoothing unit 15 performs hole filling processing by converting a region having a predetermined area or less into color data of a peripheral region. Further, the smoothed difference image data is output to the binarization unit 16.

  The binarization unit 16 captures the external setting value output from the external setting value capturing unit 11 and the smoothed difference image data output from the smoothing unit 15. Also, the binarization unit 16 reads a threshold setting method, for example, a threshold setting method proposed by Noriyuki Otsu in 1979 (hereinafter referred to as Otsu's threshold setting method) from the acquired external setting values. Further, the binarization unit 16 binarizes the acquired smoothed difference image data using a threshold setting method instructed by an external setting value. Further, the binarization unit 16 outputs the binarized difference image data to the contour extraction unit 17. For example, the binarization unit 16 binarizes the difference image data 100 as illustrated in FIG. 3A based on the threshold value, and generates the binarized difference image data 110 as illustrated in FIG. FIG. 3 is a diagram for explaining processing of input image data. As shown in FIG. 3A, the input image data 100 is composed of, for example, a background 101 and a target object 102, and the target object 102 is, for example, holes 103 and 104 penetrating the target object 102. have.

  The contour extraction unit 17 takes in the binarized difference image data output from the binarization unit 16, and from the binarized difference image data that has been taken in, the contour of the outermost periphery of the region with the largest area (hereinafter referred to as “outer”) , Referred to as outer contour data). The contour extracting unit 17 outputs the extracted outer contour data to the centroid calculating unit 18, the inner contour extracting unit 19, and the polar coordinate value generating unit 20. For example, the contour extracting unit 17 extracts the outer contour data 121 as shown in FIG. 3C from the binarized difference image data 110 as shown in FIG. That is, in the present embodiment, the feature points of the target object are not extracted as in the prior art, but the target object is extracted from the input image data by background difference processing, and the extracted target object is binarized and further binarized. The contour is extracted from the converted image data.

  The center-of-gravity calculation unit 18 takes in the outer contour data output from the contour extraction unit 17, calculates the center-of-gravity position of the contour using the acquired outer contour data, and outputs the calculated center-of-gravity position data to the polar coordinate value generation unit 20. .

  The inner contour extracting unit 19 takes in the binarized difference image data output from the binarizing unit 16 and the outer contour data output from the contour extracting unit 17, and black pixels having a predetermined area or more in the taken out outer contour. A portion is extracted, and contour data (hereinafter referred to as inner contour data) of the extracted black pixel portion is extracted. For example, as shown in FIGS. 3B and 3C, the inner contour extracting unit 19 extracts black pixel portions 111 and 112 having a predetermined area or more in the outer contour 121 extracted by the contour extracting unit 17. Then, the inner contour data 131 and 132 of the extracted black pixel portion are extracted. Further, the inner contour extraction unit 19 outputs the extracted inner contour data to the polar coordinate value generation unit 20.

  The polar coordinate value generation unit 20 includes an external setting value output from the external setting value capturing unit 11, outer contour data output from the contour extraction unit 17, centroid position data output from the centroid calculation unit 18, and an inner contour extraction unit 19. Captures the inner contour data output by. In addition, the polar coordinate value generation unit 20 extracts an angular interval for generating a polar coordinate value from the acquired external setting value. Further, the polar coordinate value generation unit 20 generates the polar coordinate value by converting the captured center-of-gravity position data and contour data into polar coordinate representation for each angle based on the external setting value, and identifies the polar coordinate value of the generated outer contour data as the identification unit To 21. Furthermore, the polar coordinate value generation unit 20 converts the captured gravity center position data and inner contour data into polar coordinate representations for each angle based on the external setting value to generate polar coordinate values, and identifies the polar coordinate values of the generated inner contour data To the unit 21.

  A method for generating polar coordinate values will be described with reference to an example of a method for generating polar coordinate values in FIG. The case where the outer contour 201, the inner contour 202, and the gravity center position O are extracted from the input image data as shown in FIG. 4A will be described. The polar coordinate value generation unit 20 has coordinates P1 (x1, y1) on the contour 201 corresponding to the angle θ formed with the horizontal axis x and θ2 = θ + Δ degrees (Δ degrees is a predetermined angular interval, for example, an interval of 5 degrees). The coordinates P2 (x2, y2) on the contour 201 corresponding to are extracted. Then, the polar coordinate value generation unit 20 calculates a distance L1 between the extracted feature point P1 (x1, y1) and the centroid position O and a distance L2 between the feature point P2 (x2, y2) and the centroid position O. Furthermore, the polar coordinate value generation unit 20 calculates the average of the distance L1 and the distance L2, and sets the calculated average value of the distances as the distance L at the angles θ1 to θ2. In this process, the coordinates on the contour 201 as shown in FIG. 4B are replaced with polar coordinate expressions having a relationship between distance and angle as shown in FIG.

  Returning to FIG. 1, the identification unit 21 receives the polar coordinate value of the outer contour data and the polar coordinate value of the inner contour data output from the polar coordinate value generation unit 20, and receives the polar coordinate value of the received outer contour data and the polar coordinate of the inner contour data. It is determined whether or not the value is written and stored in the polar coordinate value storage unit 22. If it is determined that the data has not been written, the identification unit 21 identifies the object included in the input image data as an unregistered object, and associates the polar coordinate values of the outer contour data and the inner contour data with the polar coordinate value storage unit 22. To write and memorize. If it is determined that the data has been written, the identification unit 21 identifies the target object included in the input image data as a registered target object. Further, the identification unit 21 outputs the identification result to the identification result output unit 23.

  The polar coordinate value storage unit 22 stores the polar coordinate value of the outer contour data and the polar coordinate value of the inner contour data. FIG. 5 is an example of a configuration of polar coordinate values stored in the polar coordinate value storage unit 22. As shown in FIG. 5, the polar coordinate value storage unit 22 has outer contour data and inner contour data, and the outer contour data and the inner contour data associate the distance from the center of gravity position to the contour coordinates for each predetermined angle. And remember. Further, a combination of these data is stored in association with each object.

The identification result output unit 23 captures the identification result output from the identification unit 21 and, based on the captured identification result, for example, “the object included in the input image data matches the already registered data” or “ An identification result such as “the object included in the input image data does not match the registered data” is output to the image display device 3.
Further, the image display device 3 displays the identification result output from the identification result output unit 23 on the display unit.

  Next, an object registration procedure will be described with reference to the flowchart of FIG. First, the external setting value reading unit 11 takes in external setting values that are various settings used in the object identification process (step S1). The external set value reading unit 11 outputs the acquired external set value to the background difference processing unit 14, the smoothing unit 15, the binarization unit 16, and the polar coordinate value generation unit 20.

Next, the image data capturing unit 12 captures background image data that does not include an object captured by the camera 2 (step S2). The image data capturing unit 12 writes and stores the captured background image data in the background image storage unit 13, and outputs the captured background image data to the background difference processing unit 14.
Next, the image data capturing unit 12 captures input image data captured by the camera 2 (step S3). Further, the image data capturing unit 12 outputs the captured input image data to the background difference processing unit 14.

Next, the background difference processing unit 14 reads the background image data stored in the background image storage unit 13, the external setting value output from the external setting value capturing unit 11, and the input image data output from the image data capturing unit 12. And capture. Next, the background difference processing unit 14 reads out various parameters related to the background difference processing from the fetched external setting values. Next, the background difference processing unit 14 calculates the difference between the background image data and the input image data using the various parameters relating to the read background difference processing, and outputs the calculated difference image data to the smoothing unit 15 (step). S4).
Next, the smoothing unit 15 captures the external setting value output from the external setting value capturing unit 11 and the difference image data output from the background difference processing unit 14. Next, the smoothing unit 15 smoothes the difference image data based on the acquired external setting value, and outputs the smoothed difference image data to the binarization unit 16 (step S5).
Next, the binarization unit 16 captures the external setting value output from the external setting value capturing unit 11 and the smoothed difference image data output from the background difference processing unit 14. Also, the binarization unit 16 binarizes the captured smoothed difference image data using an external setting value threshold setting method (for example, Otsu's threshold setting method) and binarizes the difference image data. Is output to the contour extraction unit 17 (step S6). For example, in the binarized difference image data, as shown in FIG. 3B, the outside of the contour is a black pixel, and the inside of the contour that is the object is a white pixel. In addition, regions 111 and 112 having an inner contour in the contour are black pixels.

Next, the contour extraction unit 17 takes in the binarized difference image data output from the binarization unit 16, extracts the outer contour data from the binarized difference image data, and extracts the extracted outer contour The data is output to the center of gravity calculation unit 18 and the inner contour extraction unit 19 (step S7).
Next, the center-of-gravity calculation unit 18 takes in the outer contour data output from the contour extraction unit 17, calculates the center-of-gravity position of the outer contour using the acquired outer contour data, and uses the calculated center-of-gravity position data as the polar coordinate value generation unit 20 (Step S8).
Next, the polar coordinate value generation unit 20 captures the external setting value output from the external setting value capturing unit 11, the outer contour data output from the contour extraction unit 17, and the centroid position data output from the centroid calculation unit 18. Next, the polar coordinate value generation unit 20 calculates the average value of the distance from the centroid position to the outer contour for each angle according to the external setting value, and calculates the angle and distance from the captured centroid position data and outer contour data. The polar coordinate value is generated by converting into the polar coordinate expression used, and the generated polar coordinate value is output to the identification unit 21 (step S9).

  A polar coordinate value generation procedure performed by the polar coordinate value generation unit 20 will be described with reference to the flowchart of FIG. The polar coordinate value generation unit 20 performs all the processes for each predetermined angle (for example, 5 degrees) extracted from the external setting value, from 0 degrees to 360 degrees (steps S101 to S105). First, the polar coordinate value generation unit 20 calculates the distance L1 from the position of the center of gravity at the angle of 0 degrees to the coordinates of the outer contour data (step S102). Next, the polar coordinate value generation unit 20 calculates a distance L2 from the position of the center of gravity at the angle of 5 degrees to the coordinates of the outer contour (step S103). Next, the polar coordinate value generation unit 20 calculates an average value of the distance L1 and the distance L2 as a distance with an angle of 0 degrees to 5 degrees (step S104). Next, the average value L of the distance between the angle 5 degrees and the angle 10 degrees is calculated, and these processes are performed from the angle 0 degrees to 360 degrees.

Next, the inner contour extraction unit 19 takes in the binarized difference image data output from the binarization unit 16 and the contour data and outer contour data output from the contour extraction unit 17. Next, the inner contour extraction unit 19 searches for a black pixel region having a predetermined area or more in the outer contour data using the binarized difference image data and the outer contour data (step S10).
Next, as a result of the search, the inner contour extraction unit 19 determines whether or not a black pixel region having a predetermined area exists in the outer contour data (step S11).
As a result of the determination, when a black pixel region having a predetermined area exists in the outer contour data (step S11; Yes), the inner contour extraction unit 19 extracts the inner contour data of the black pixel region and extracts the extracted inner contour. Data is output to the polar coordinate value generation unit 20. Specifically, the inner contour extracting unit 19 compares the difference image data 110 that is binarized to determine whether there is a black area larger than a predetermined area inside the outer contour data 121, and based on the comparison result, the region 111. And 112 are extracted. Then, the inner contour extracting unit 19 extracts the extracted inner contour data 131 and 132 of the black pixel region and outputs them to the extracted polar coordinate value generating unit 20.
If the result of determination is that there is no region of a predetermined black area in the outer contour data (step S11; No), the process proceeds to step S14.

Next, the polar coordinate value generation unit 20 takes in the centroid position data output from the centroid calculation unit 18 and the inner contour data output from the inner contour extraction unit 19. Next, the polar coordinate value generation unit 20 generates polar coordinate values by converting the inner contour data into polar coordinate representations using the acquired center-of-gravity position data and inner contour data in the same manner as the polar coordinate value generation of the outer contour data. The polar coordinate value thus output is output to the identification unit 21 (step S12).
Next, the inner contour extraction unit 19 determines whether or not the polar coordinate values of all the inner contour data in the outer contour data have been generated (step S13).
When generation of polar coordinate values of all inner contour data has not been completed (step S13; No), the inner contour extraction unit 19 and polar coordinate value generation unit 20 repeat steps S10 to S13, and all the outer contour data in the outer contour data Are extracted, polar coordinate values of all the extracted inner contour data are generated, and the polar coordinate values of the generated inner contour data are output to the identification unit 21.
Next, the identification unit 21 takes in the polar coordinate values of the outer contour data and the inner contour data output from the polar coordinate value generation unit 20, and the object data of the combination of the fetched polar coordinate values is already stored in the polar coordinate value storage unit 22. It is determined whether or not it has been done. When a new object is registered, since it is not yet stored in the polar coordinate value storage unit 22, the identification unit 21 associates, writes and stores the polar coordinate value of the captured outer contour data and the polar coordinate value of the inner contour data. (Step S14).
This completes the registration procedure of the object.

Next, a procedure for identifying an object will be described with reference to the flowchart of FIG. First, an input image taken by the camera 2 is captured, and steps S1 to S13 are performed in the same manner as the registration procedure to generate outer contour data and inner contour data (step S201).
Next, the identification unit 21 compares the polar coordinate value of the acquired outer contour data with the polar coordinate value of the outer contour data of the object stored in the polar coordinate value storage unit 22 (step S202).
The comparison method is performed as follows, for example. The absolute value of the difference between the distance L _m of 0 to π / 36 stored in the polar storage unit 22 and the distance L _i of 0 to π / 36 generated from the captured input image data is calculated. Next, the absolute value of the difference between the distance L _{m + 1} of π / 36 to 2π / 36 stored in the polar storage unit 22 and the distance L _{i + 1} of π / 36 to 2π / 36 generated from the captured input image data is obtained. calculate. In this way, the absolute value of the difference is calculated for all angles, and the sum of the absolute values of the calculated differences is calculated. Next, calculate the distance _{L m + 1} of π / 36~2π / 36 stored in the polar coordinate storage section 22, the absolute value of the difference between the distance _{L i} of 0~π / 36 generated from the input image data taken in . This procedure is repeated, and the absolute value of the difference between the distance L _m of 0 to π / 36 stored in the polar coordinate storage unit 22 and the distance L _{i + 36} of 35π / 36 to 0 generated from the captured input image data is calculated. . That is, the absolute value of the difference is calculated by shifting the distance stored in the polar coordinate value storage unit 22 by π / 36, and the sum of the absolute values of the difference is similarly calculated.
In this way, the identification unit 21 calculates the sum of absolute values of differences shifted by π / 36, calculates the shift amount (angle) with the smallest sum of absolute values of differences, and calculates the smallest absolute value. The sum is extracted as a comparison value.

Next, the identification unit 21 compares the extracted comparison value with a predetermined threshold value. If the comparison value is smaller than the predetermined comparison value, the identification unit 21 includes the object contour data stored in the polar coordinate storage unit 22 and the input image data. It is determined whether or not the outer contour data of the target object matches (step S203).
When the outer contour data of the target object stored in the polar coordinate storage unit 22 does not match the outer contour data of the target object included in the input image data (step S203; No), the identification unit 21 determines that the target object is one The identification result is output to the identification result output unit 23, and the generated outer contour data and inner contour data are associated with each other and written and stored in the polar coordinate value storage unit 22 (step S205). The identification result output unit 23 captures the identification result output from the identification unit 21 and outputs the identification result based on the captured identification result to the image display device 3. Further, the image display device 3 displays the identification result output from the identification result output unit 23 on the display unit.

A specific example of identification in steps S202 and S203 will be described with reference to the identification method in FIG. In order to simplify the description, a case will be described in which polar coordinate values are generated in 16 equal parts (22.5 degree intervals). When the polar coordinate value stored in the polar coordinate storage unit 22 is FIG. 9A and the polar coordinate value generated from the input image data is FIG. 9B, first, the distance L _m = 6 and the distance L between the angles 0 to θ1. Calculate the absolute value of the difference from _i = 3 = 3. This procedure is repeated from the angle θ15 to 0 to calculate the sum of absolute values of differences between all angles = 18. Next, the polar coordinate values stored in the polar coordinate storage unit 22 are compared by shifting the angle of θ1. That is, the absolute value L = 4 of the difference between the distance L _{m + 1} = 7 of the angles θ1 to θ2 and the distance L _i = 3 is calculated. This procedure is repeated from angle θ15 to 0 to calculate the sum of absolute values of all angle differences = 0. In this way, the angle θ1 is shifted and compared, and θ1 having the smallest sum of absolute values of all angle differences is extracted as a comparison value. Then, the sum of the absolute values of the extracted differences = 0 is compared with a predetermined threshold value to determine that it is smaller than the predetermined comparison value, and the outer contour data of the object stored in the polar coordinate storage unit 22; It discriminate | determines that the outer outline data of the target object contained in input image data correspond. If it is determined that the objects match, the comparison value θ1 is also the rotation angle of the object.

  Returning to FIG. 8, when the contour data of the target object stored in the polar coordinate storage unit 22 matches the contour data of the target object included in the input image data (step S203; Yes), the identification unit 21 It is determined whether or not there is inner contour data in the input image data (step S206). When there is no inner contour data in the input image data (step S206; No), the identification unit 21 determines whether the inner contour data is stored in association with the outer contour data in the polar coordinate storage unit 22 (step S207). . When the inner contour data is stored in the polar coordinate storage unit 22 in association with the outer contour data (step S207; Yes), the process proceeds to step S204, and the inner contour data is stored in the polar coordinate storage unit 22 in association with the outer contour data. If yes (step S207; No), the process proceeds to step S213.

When the input image data includes inner contour data (step S206; Yes), the identification unit 21 calculates the area of the inner contour data generated from the input image data. The contour data is arranged (step S208).
Next, the identifying unit 21 selects the inner contour data having the largest area (step S209), and is associated with the selected inner contour data and the outer contour data identified as matching in steps S202 to S203, and is polar coordinates. Whether the inner contour data stored in the storage unit 22 matches or not is compared by the same method as the outer contour data comparison method (step S210).

As a result of the comparison, when the inner contour data of the input image data and the inner contour data stored in the polar coordinate storage unit 22 do not match (step S211; No), the process proceeds to step S204, and the object does not match. Process. As a result of the comparison, when the inner contour data of the input image data matches the inner contour data stored in the polar coordinate storage unit 22 (step S211; Yes), all the inner contour data generated from the input image data Then, it is determined whether or not the comparison has been completed for all the inner contour data stored in the polar coordinate storage unit 22 in association with the outer contour data identified as coincident (step S212).
If all the inner contour data comparisons have not been completed (step S212; No), steps S209 to S212 are repeated. When all the inner contour data comparisons are completed and the inner contour data match (step S212; Yes), it is determined that the objects match, and the identification result is output to the identification result output unit 23 (step S213). Further, the image display device 3 displays the identification result output from the identification result output unit 23 on the display unit.
This completes the object identification procedure.

  As described above, the object is extracted by the background difference process, and the contour data is extracted from the image data obtained by binarizing the extracted difference image. Then, the position of the center of gravity is calculated using the extracted contour data, converted into a polar coordinate expression using the position of the center of gravity and the contour data, and the polar coordinate value is generated, and the object is identified using the generated polar coordinate value. Therefore, high recognition accuracy can be obtained even in an image in which feature points are difficult to extract. Further, the black pixel region when binarized into the in-contour data of the object is extracted, and the contour data of the extracted black pixel region is further extracted. Then, the extracted contour data is converted into a polar coordinate expression using the gravity center position and the contour data to generate a polar coordinate value, and the object is identified using the generated polar coordinate value. Compared with the conventional method, the amount of calculation can be reduced for the target object, and for example, a mobile terminal or the like can be implemented.

[Second Embodiment]
The configuration of the object identification device according to the present embodiment will be described with reference to a block diagram illustrating an example of the configuration of the object identification device 300 in FIG. The same functional units as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted. The difference from the first embodiment is that an external set value capturing unit 301, a smoothing unit 302, a color space separating unit 303, a binarizing unit 304, a contour extracting unit 305, a centroid calculating unit 306, A channel selection unit 307 and a polar coordinate value storage unit 310.

  The external setting value capturing unit 301 captures external setting values, which are various settings used for object identification processing, from the outside. The external setting value capturing unit 301 receives the external setting values from the background difference processing unit 14, the smoothing unit 302, the color space separation unit 303, and the like. The data is output to the value conversion unit 304, the channel selection unit 307, the polar coordinate value generation unit 308, and the identification unit 309.

  The color space separation unit 303 captures the external setting value output from the external setting value capturing unit 301 and the smoothed difference image data output from the smoothing unit 302. Further, the color space separation unit 303 separates the smoothed difference image data into color spaces such as RGB and HSV (hue, saturation, brightness) according to the external setting value, and separates the color space images of each channel. Data is output to binarization 304. FIG. 11 is a diagram for explaining color space channel decomposition of RGB image data. When the external setting value is set to separate the RGB image data into R, G, and B channels, the color space separation unit 303 separates the difference image data into the image data of each channel as shown in FIG.

  The binarization unit 304 outputs the external setting value output from the external setting value capturing unit 301, the smoothed difference image data (RGB image data) output from the smoothing unit 302, and the color space separation unit 303. Capture color space image data for a channel. Also, the binarization unit 304 binarizes the smoothed difference image data and the color space image data of each channel according to the external setting value, and binarizes the smoothed difference image data. The color space image data of each channel is output to the contour extraction unit 305.

  The contour extraction unit 305 captures the binarized difference image data output from the binarization unit 304 and the binarized color space image data of each channel, and the captured difference image data and the color of each channel Outer contour data is extracted from each of the spatial image data. In addition, the contour extraction unit 305 outputs the outer contour data of the extracted difference image data to the centroid calculation unit 306, and outputs each outer contour data of the extracted color space image data of each channel to the polar coordinate value generation unit 308.

  The center-of-gravity calculation unit 306 takes in the outer contour data of the difference image data output from the contour extraction unit 305, calculates the center of gravity position from the outer contour data of the acquired difference image data, and uses the calculated center-of-gravity position data as the polar coordinate value generation unit 308. Output to.

The channel selection unit 307 captures the external setting value output from the external setting value capturing unit 301, and extracts color space separation instruction information indicating which color space separation method to use from the captured external setting value. In addition, the channel selection unit 307 generates channel selection data that switches the polar coordinate value generation unit 308 from time to time using the outer contour data of which channel based on the extracted color space separation instruction information. The selected channel selection data is output to the polar coordinate value generation unit 308.
For example, when the external setting values are separated into RGB, the channel selection unit 307 first outputs channel selection data for generating polar coordinate values using the outer contour data of the RGB image data to the polar coordinate value generation unit 308. The channel selection unit 307 captures the generation completion data output from the polar coordinate value generation unit 308, selects the R channel that is the channel of the color space that has not yet generated the polar coordinate values after capturing the generation completion data, and selects the selected R Channel selection data for generating polar coordinate values using the outer contour data of the image data is output to the polar coordinate value generation unit 308. In this manner, channel selection data for generating polar coordinate values from the respective outer contour data of all color space image data is sequentially output to the polar coordinate value generating unit 308.

  The polar coordinate value generation unit 308 includes an external setting value output from the external setting value capturing unit 301, outer contour data of difference image data output from the contour extraction unit 305, outer contour data of each color space, and a centroid calculating unit 306. The center-of-gravity position data to be output and the channel selection data output from the channel selection unit 307 are captured. Further, the polar coordinate value generation unit 308 generates a polar coordinate value according to the externally set value for each of the captured outer contour data according to the channel selection data. Further, the polar coordinate value generation unit 308 generates the generation completion data after the generation of each polar coordinate value is completed, and outputs the generated generation completion data to the channel selection unit 307.

The identification unit 309 includes an external setting value output from the external setting value capturing unit 301, a polar coordinate value of the outer contour data of the difference image data output from the polar coordinate value generation unit 308, and a polar coordinate value of the outer contour data of each color space image data. And capture. Also, the identification unit 309 extracts channel priority data from the acquired external setting value. Further, the identification unit 309 first determines whether or not the combination of each polar coordinate value acquired matches the polar coordinate value stored in the polar coordinate value storage unit 310 with respect to the polar coordinate value generated from the RGB image data. Further, the identification unit 309 determines whether or not the polar coordinate value generated from the color space image data corresponding to the channel priority data matches the polar coordinate value stored in the polar coordinate value storage unit 310.
Further, the identification unit 309 identifies the object based on the determination result, and outputs the identification result to the identification result output unit 23. Further, in the case of a polar coordinate value that is not stored in the polar coordinate value storage unit 310 as a result of the determination, the identification unit 309 writes and stores each polar coordinate value in the polar coordinate value storage unit 310.

  The polar coordinate value storage unit 310 stores the polar coordinate value of the outer contour data of the difference image data and the polar coordinate value of the outer contour data of each color space image data in association with each predetermined angle. FIG. 12 is an example of the configuration of polar coordinate values stored in the polar coordinate value storage unit 310. As shown in FIG. 12, the polar coordinate value storage unit 310 includes RGB image data and outer contour data of image data of each color space channel (R, G, B), and each outer contour data is, for example, 5 degrees for each angle. The distance from the center of gravity position to the feature point of the contour is stored in association with each other. Further, a combination of these data is stored in association with each object.

Next, the registration procedure of the target object in 2nd embodiment is demonstrated using the flowchart of FIG. About the same process as 1st embodiment, description is abbreviate | omitted using the same code | symbol. First, steps S1 to S5 are performed in the same manner as in the first embodiment.
Next, the color space separation unit 303 captures the external setting value output from the external setting value capturing unit 301 and the smoothed difference image data output from the smoothing unit 302. Next, the color space separation unit 303 separates the difference image data smoothed based on the external setting values into R, G, and B color space components, and binarizes the separated color space image data of each channel 304. Output to. (Step S301).
Next, the binarization unit 304 outputs the external setting value output from the external setting value capturing unit 301, the smoothed difference image data output from the smoothing unit 302, and each channel output from the color space separation unit 303. The color space image data is captured. Also, the binarization unit 304 uses the externally set value threshold setting method (for example, Otsu's threshold setting method) to convert the smoothed difference image data and the color space image data of each channel. The binarized difference image data and the color space image data of each channel are output to the contour extraction unit 305 (step S302).

Next, the contour extraction unit 305 takes in the binarized difference image data output from the binarization unit 304, extracts outer contour data from the binarized difference image data, and extracts the extracted outer contour Data is output to the gravity center calculation unit 18 (step S303).
Next, the centroid calculating unit 18 takes in the outer contour data output from the contour extracting unit 305, calculates the centroid position of the outer contour using the acquired outer contour data, and uses the calculated centroid position data as the polar coordinate value generating unit 308. (Step S304).

Next, the channel selection unit 307 captures the external setting value output by the external setting value capturing unit 301, and extracts color space separation instruction information from the captured external setting value. Further, the channel selection unit 307 generates channel selection data for selecting RGB image data based on the extracted color space separation instruction information, and outputs the generated channel selection data to the polar coordinate value generation unit 308.
Next, the polar coordinate value generation unit 308 outputs the external setting value output from the external setting value capturing unit 301, the outer contour data of the difference image data output from the contour extraction unit 17, and the centroid position data output from the centroid calculation unit 18. And the channel selection data output by the channel selection unit 307 are captured.
Next, the polar coordinate value generation unit 308 generates polar coordinate values of the difference image data of the RGB image data according to the channel selection data. Next, the polar coordinate value generation unit 308 converts the captured center-of-gravity position data and outer contour data into the polar coordinate expression by converting the average value of the distance from the center of gravity position to the outer contour for each angle according to the external setting value. A value is generated, and the generated polar coordinate value is output to the identification unit 309 (step S305).

Next, the channel selection unit 307 outputs channel selection data for selecting the R channel according to the external setting value to the polar coordinate value generation unit 308 (step S306).
Next, the polar coordinate value generation unit 308 outputs the external setting value output from the external setting value capturing unit 301, the outer contour data of the color space image data of each channel output from the contour extraction unit 17, and the centroid calculation unit 18 The center-of-gravity position data and the channel selection data output by the channel selection unit 307 are captured.
Next, the polar coordinate value generation unit 308 generates polar coordinate values of the R channel image data according to the channel selection data. Next, the polar coordinate value generation unit 308 converts the captured center-of-gravity position data and outer contour data into the polar coordinate expression by converting the average value of the distance from the center of gravity position to the outer contour for each angle according to the external setting value. A value is generated, and the generated polar coordinate value is output to the identification unit 309 (step S307).
Next, the polar coordinate value generation unit 308 determines whether or not the generation of polar coordinate values for all color space channels (R, G, B) has been completed (step S308). When the polar coordinate value generation of all channels is not completed (step S308; No), step S306 to step S308 are repeated. When the polar coordinate value generation of all the channels is completed (step S308; Yes), the polar coordinate value generation unit 308 identifies each polar coordinate value of each contour data in the generated RGB image data and the color space image data of each color channel. Output to.

Next, the identification unit 309 captures each polar coordinate value of each outer contour data output from the polar coordinate value generation unit 308, and whether or not the object data of the combination of the captured polar coordinate values is already stored in the polar coordinate value storage unit 310. Determine whether. When a new object is registered, since it is not yet stored in the polar coordinate value storage unit 310, the identification unit 309 associates, writes, and stores the polar coordinate values of each piece of outer contour data acquired (step S309).
This completes the registration procedure of the object.

Next, a procedure for identifying an object in the second embodiment will be described with reference to the flowchart of FIG. First, an input image photographed by the camera 2 is captured, and steps S1 to S5 and steps S301 to S308 are performed in the same manner as the registration procedure to generate each polar coordinate value of each outer contour data (step S401).
Next, the identification unit 309 compares the polar coordinate value of the outer contour data generated from the RGB image data with the outer contour data generated from the RGB image data stored in the polar coordinate value storage unit 310 (step S402). ). The polar coordinate value comparison method is performed in the same manner as in the first embodiment.

  Next, when the outer contour data of the target object stored in the polar coordinate storage unit 310 and the outer contour data of the target object included in the input image data do not match (step S403; No), the identifying unit 309 It is identified that the objects do not match, the identification result is output to the identification result output unit 23, and the generated outer contour data is associated with each other and written and stored in the polar coordinate value storage unit 310 (step S405). Further, the identification result output unit 23 captures the identification result output by the identification unit 309 and outputs the identification result based on the captured identification result to the image display device 3. Further, the image display device 3 displays the identification result output from the identification result output unit 23 on the display unit.

Next, when the outer contour data of the target object stored in the polar coordinate storage unit 310 matches the outer contour data of the target object included in the input image data (step S403; Yes), the identifying unit 309 The external setting value output by the capturing unit 11 is captured, and channel priority data is extracted from the captured external setting value. Next, the identification unit 309 selects a color space channel corresponding to the channel priority data (step S406).
Next, the identification unit 309 uses the polar coordinate value of the selected color space channel and the polar coordinate value of the same color space channel stored in the polar coordinate value storage unit 310 as the polar coordinate value comparison method of the first embodiment. The same is done (step S407).

Next, it is determined whether or not the outer contour data of the object color space channel stored in the polar coordinate storage unit 310 matches the outer contour data of the object color space channel included in the input image data. (Step S408).
As a result of the determination, if they do not match (step S408; No), the identification unit 309 identifies that the objects do not match, and performs steps S404 to S405.
As a result of the determination, if they match (step S408; Yes), the identification unit 309 determines whether the determination has been completed for all color space channels (step S409).
If the determination has not been completed for all the color space channels (step S409; No), steps S406 to S409 are repeated. When the determination is completed for all the color space channels (step S409; Yes), the identification unit 309 determines that the objects match and outputs the identification result to the identification result output unit 23 (step S410). Further, the image display device 3 displays the identification result output from the identification result output unit 23 on the display unit.
This completes the object identification procedure.

  As described above, in addition to the first embodiment, the polar coordinate value of the outer contour data of the object is generated based on the image data obtained by separating the RGB image data of the input image data into the color space, and the generated polar coordinate value of each channel Also used to identify objects. As a result, the component separated into each channel differs depending on the color space component of the difference image data. For example, an object having a specific color space component such as blue uses contour data generated from the RGB difference image data. In addition, the contour data generated from the differential image data of only the B channel can be used for more accurate identification. Furthermore, even when the object has a pattern or pattern, the pattern or pattern can be extracted with high accuracy by performing color space separation. The object can be identified with high accuracy.

  In the second embodiment, the example in which the color space components are separated and identified has been described. However, the inner contour data of the first embodiment may be extracted. In this case, in addition to the configuration of the second embodiment, an inner contour extraction unit may be provided. In the second embodiment, the example in which the color space is separated into R, G, and B has been described. However, the same effect can be obtained even in a color space such as HSV.

  In the first embodiment, the example in which the polar coordinate values of the inner contour data are compared in the descending order of the area of the inner contour data has been described, but the order may be random or a predetermined order. Furthermore, in the first embodiment, the example in which the polar coordinate values of all the inner contour data are compared has been described, but the comparison of the polar coordinate values of the inner contour data having a predetermined area or less is performed for comparison. Accordingly, the identification may be omitted. Alternatively, the comparison of a predetermined area or less may be omitted if polar coordinate values of a predetermined number or more of inner contour data match.

  In this embodiment, for example, the average value of the distance L1 between the coordinates of the contour data corresponding to the angle θ degrees and the center of gravity and the distance L2 between the coordinates of the contour data corresponding to the angle θ + Δ degrees and the center of gravity is calculated. Although the example of generating the polar coordinate value has been described, the distance L1 between the coordinates of the contour data corresponding to the angle θ degree and the center of gravity may be the polar coordinate value. In addition, when the object is configured by a close outer contour and an inner contour such as a donut shape, for example, an average value of the distance L1 between the outer contour and the center of gravity and the distance L2 between the inner contour and the center of gravity. The polar coordinate value may be generated using as a distance.

  In this embodiment, the example in which the number of times of smoothing by the smoothing unit 15 or 302 is specified by the external setting value captured by the external setting value capturing unit 11 or 301 has been described. Alternatively, the number of times may be set according to the input image or the object. In this case, for example, the complexity of the input image may be determined based on the magnitude of the high-frequency component, the distribution degree of the component in the histogram, and the like, and the number of times of smoothing may be increased for complex image data. Further, as a result of the calculation as described above, when the number of times of averaging is 1, for example, color space separation may be performed without performing the averaging process.

  Further, in the present embodiment, an example has been described in which the distance from the center of gravity position to the coordinates of the outer contour data or the coordinates of the inner contour data is written and stored in the polar coordinate value storage unit 22 or 310 as they are, but all polar coordinate values have been calculated. Thereafter, the maximum value of the distance may be extracted, and the extracted maximum value may be normalized as 1, for example, and written and stored in the polar coordinate value storage unit 22 or 310. Alternatively, the comparison may be performed by normalizing the polar coordinate values.

  Moreover, although the binarization part 16 or 304 demonstrated the example set with the threshold selection method of Otsu in this embodiment, you may use another general threshold selection method.

FIG. 15 is a diagram illustrating a usage example of the object identification device 1 or 300 according to the present embodiment. The object identification device 1 or 300 according to the present embodiment can be used as an intergenerational communication tool using a memorable item with a clogged image (associated). Specifically, the grandchild who is the data sender puts memorable items such as shells picked up on the trip on the imaging device in front of the TV and performs simple operations on the registration device, so You can edit the video you want to stuff (or link). Then, the grandparents who have received the memorabilia simply place the memorabilia on the imaging device in front of the television, and the reading device operates to view the video associated with the memorabilia. Can do. In such a system, it is necessary to register and identify the memorable item in the registration device and the reading device. These processes are assumed to be performed by a personal computer or a mobile terminal in a general home. For this reason, the amount of calculation is small, and furthermore, high-precision identification accuracy is required. Even in such a system, by applying the invention of the present embodiment, the amount of calculation is small, and furthermore, highly accurate identification accuracy can be performed.
As described above, the object identification device 1 or 300 according to the present embodiment operates by a simple and intuitive operation of placing memories, so that even users with low IT literacy can share data via the network. Can do.

It should be noted that a program for realizing the functions of the respective units in FIG. 1 and FIG. 10 of the embodiment is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. You may process each part by. Here, the “computer system” includes an OS and hardware such as peripheral devices.
Further, the “computer system” includes a homepage providing environment (or display environment) if the WWW system is used.
“Computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM (Read Only Memory), a portable medium such as a CD-ROM, and a USB (Universal Serial Bus) I / F (interface). A storage device such as a USB memory or a hard disk built in a computer system. Furthermore, a “computer-readable recording medium” dynamically holds a program for a short time, like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. It also includes those that hold programs for a certain period of time, such as volatile memory inside computer systems that serve as servers and clients in that case. The program may be a program for realizing a part of the above-described functions, or may be a program that can realize the above-described functions in combination with a program already recorded in a computer system.

DESCRIPTION OF SYMBOLS 1 ... Object identification apparatus 2 ... Camera 3 ... Image display apparatus 11 ... External setting value acquisition part 12 ... Image data acquisition part 13 ... Background image storage part 14 ... Background difference Processing unit 15 ... smoothing unit 16 ... binarization unit 17 ... contour extraction unit 18 ... centroid calculation unit 19 ... inner contour extraction unit 20 ... polar coordinate value generation unit 21 ... Identification unit 22 ... polar coordinate value storage unit 23 ... identification result output unit

Claims (5)

A background difference processing unit that extracts a figure included in the input image by a background difference method from the input image;
A binarization unit that binarizes the graphic image data extracted by the background difference processing unit;
A contour extraction unit that extracts contour data from image data of the figure binarized by the binarization unit;
A centroid calculating unit for calculating a centroid position of the contour data extracted by the contour extracting unit;
An inner contour extraction unit that extracts a black pixel region that is inside the contour data extracted by the contour extraction unit and has a predetermined area or more, and extracts inner contour data of the extracted black pixel region;
Contour data among extracted by the said contour extraction unit from the center of gravity position distance and the centroid calculating section from the center of gravity position where the center of gravity calculation unit has calculated to the outermost portion of the contour data, wherein the contour extraction unit has extracted was calculated A polar coordinate value generation unit that calculates a distance to the outermost peripheral part for each predetermined angle, converts the contour data and the inner contour data into a polar coordinate expression, and generates respective polar coordinate values;
A polar coordinate value storage unit for storing the polar coordinate value of the contour data of the figure to be identified and the polar coordinate value of the inner contour data ;
And polar value of the contour data stored before Symbol polar coordinate value storage unit, and a Gokuza target value of the contour data, wherein the polar coordinate value generating unit has generated compared while shifting by a predetermined angle, to the polar coordinate value storage unit By comparing the polar coordinate value of the stored inner contour data with the polar coordinate value of the inner contour data generated by the polar coordinate value generation unit while shifting by a predetermined angle, it is possible to identify the figure included in the input image. An identification unit to perform ,
Object identification device, characterized in that it comprises a. A color space separation unit for separating image data of a graphic extracted from the input image into a plurality of color space components;
Further comprising
The binarization unit binarizes each color space component separated by the color space separation unit,
The contour extraction unit extracts contour data from image data of each color space component binarized by the binarization unit,
The polar coordinate value of each contour data of each color space component stored in the polar coordinate value storage unit and the polar coordinate expression value of each contour data of each color space component generated by the polar coordinate value generation unit are compared while shifting by a predetermined angle. The object identification apparatus according to claim 1, wherein a figure included in the input image is identified. The polar coordinate value generating unit has generated to extract the maximum value of the distance among the polar coordinate values, according to claim 1 or claim 2, characterized in that normalization is performed based on the maximum value of the extracted distance Object identification device. In an object identification method of an object identification device comprising a polar coordinate value storage unit for storing polar coordinate values of contour data of a graphic to be identified and polar coordinate values of inner contour data ,
A background difference processing unit, a background difference processing step for extracting a figure included in the input image from the input image by a background difference method;
A binarization step for binarizing the graphic image data extracted by the background difference processing unit ;
A contour extraction step for extracting contour data from image data of the figure binarized by the binarization unit ;
A centroid calculating step of calculating a centroid position of the contour data extracted by the contour extracting unit ;
The inner contour extraction unit extracts a black pixel region that is inside the contour data extracted by the contour extraction unit and has a predetermined area or more, and extracts inner contour data of the extracted black pixel region. An inner contour extraction process;
Polar coordinate value generating portion, wherein the contour extraction unit from the center of gravity position distance and the centroid calculating section from the center of gravity position where the center of gravity calculation unit has calculated to the outermost portion of the contour data, wherein the contour extraction unit has extracted was calculated The polar coordinate value generation step of calculating the distance to the outermost peripheral portion of the extracted inner contour data for each predetermined angle, converting the contour data and the inner contour data into polar coordinate representations, and generating the respective polar coordinate values When,
Identification unit compares while shifting the polar coordinates of contour data stored in the polar coordinate value storage unit, and a Gokuza target value of the contour data, wherein the polar coordinate value generating unit has generated by a predetermined angle, the polar coordinate value A figure included in the input image by comparing the polar coordinate value of the inner contour data stored in the storage unit and the polar coordinate value of the inner contour data generated by the polar coordinate value generation unit while shifting the polar coordinate value by a predetermined angle. An identification process for identifying
An object identification method characterized by comprising: In the computer of the object identification device comprising a polar coordinate value storage unit for storing the polar coordinate value of the contour data of the figure to be identified and the polar coordinate value of the inner contour data ,
A background difference processing step of extracting a figure contained in the input image from the input image by a background difference method;
A binarization step of binarizing the image data of a figure that has been extracted by the background difference process,
A contour extraction step for extracting contour data from the image data of the binarized figures by the binarization process,
A centroid calculation step of calculating the center of gravity of the contour data extracted by said contour extraction step,
An inner contour extracting step of extracting a black pixel region which is present inside the contour data extracted by the contour extracting step and is a black pixel region having a predetermined area or more and extracting inner contour data of the extracted black pixel region; ,
Contour inner distance and the centroid calculating section from the center of gravity position calculated by the centroid calculating step to the outermost periphery portion of the contour data extracted by the outline extraction process is extracted by said contour extraction unit from the gravity center position calculated A polar coordinate value generation step of calculating a distance to the outermost peripheral portion of the data for each predetermined angle, converting the contour data and the inner contour data into a polar coordinate representation, and generating respective polar coordinate values ;
And polar value of the contour data stored before Symbol polar coordinate value storage unit, and a Gokuza target value of the contour data generated by the polar coordinate value generating step compares while shifting by a predetermined angle, the polar coordinate value storage unit Is compared with the polar coordinate value of the inner contour data stored in the inner coordinate data and the polar coordinate value of the inner contour data generated by the polar coordinate value generating step while shifting the predetermined contour by a predetermined angle . An identification process for identifying;
An object identification program for executing