Detection device, detection method, and detection program

Abstract

This detection device is provided with: a color channel selection unit that selects a color channel on the basis of a degree of separation calculated from each of a first color channel image and a second color channel image for an original image including a background image and a hand image; the index is used for judging the separation between the background image and the hand image; and a hand image detection unit that detects the hand image included in the original image using the color channel selected by the color channel selection unit.

Description

Technical Field

[0001] This case involves detection devices, detection methods and detection procedures. Background Art

[0002] For palm vein authentication, palm print authentication, etc., it is required to accurately detect a hand image from an original image (for example, see Patent Document 1). In addition, there is known a technology for detecting a hand image from an image obtained by binarizing an original image captured by irradiating visible light using brightness (for example, see Patent Document 2).

[0003] Patent Document 1: Japanese Patent Application Publication No. 2018-026012;

[0004] Patent document 2: Japanese Patent No. 6629150.

[0005] Non-patent document 1: T.Simon, H.Joo, I.Matthews, and Y.Sheikh, "Hand keypoint detection in single images using multiview bootstrapping," in CVPR, 2017.

[0006] However, it is difficult to accurately detect the palm area. Summary of the invention

[0007] In one aspect, an object of the present invention is to provide a detection device, a detection method, and a detection program that can improve the detection accuracy of hand images.

[0008] In one embodiment, a detection device comprises: a color channel selection unit, which selects a color channel based on a degree of separation, wherein the degree of separation is calculated based on a first color channel image and a second color channel image of an original image including a background image and a hand image, respectively, and is an indicator for determining the separability of the background image and the hand image; and a hand image detection unit, which detects the hand image included in the original image using the color channel selected by the color channel selection unit.

[0009] Can improve the detection accuracy of hand images. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 (a) is a block diagram illustrating the overall structure of the detection device, and (b) is a block diagram illustrating the hardware structure of the detection device.

[0011] Figure 2 This is a flowchart showing an example of hand image detection processing performed by the detection device.

[0012] Figure 3(a) is a diagram illustrating a key point, and (b) is a diagram illustrating a small area.

[0013] Figure 4 Yes means Figure 2 FIG. 5 is a flowchart illustrating a detailed example of step S6.

[0014] Figure 5 is an example Figure 2 Detailed diagram of steps S5 to S7.

[0015] Figure 6 This is a diagram illustrating a case where the degree of separation is large.

[0016] Figure 7 This is a diagram illustrating a case where the degree of separation is large.

[0017] Figure 8 This is a diagram illustrating a case where the degree of separation is small.

[0018] Fig. 9 This is a diagram illustrating a case where the degree of separation is small.

[0019] Fig.10 This is a flowchart showing an example of the registration process.

[0020] Fig.11 This is a flowchart showing an example of the comparison process. DETAILED DESCRIPTION

[0021] Biometric authentication is a technology that uses biometric information such as fingerprints, faces, and veins to confirm a person's identity. In biometric authentication, the biometric data for comparison acquired by the sensor is compared (checked) with the pre-registered biometric data in the scene that requires confirmation, and it is determined whether the similarity is above the threshold for determining the identity, thereby confirming the identity.

[0022] It is important to accurately detect a hand image from an original image in palm vein authentication, palm print authentication, etc. However, when shooting with a general camera such as a camera built into a portable terminal, it may be difficult to separate the hand image from the background image due to the optical characteristics of the camera, reflections due to the influence of light sources, the influence of shadows, the color and pattern of the background, etc.

[0023] Therefore, in the following embodiments, a detection device, a detection method, and a detection program are described that can improve the detection accuracy of a hand image by separating the hand image from the background image regardless of the shooting conditions.

[0024] Example 1

[0025] Figure 1 (a) is a block diagram illustrating the overall structure of the detection device 100. Figure 1As illustrated in (a), the detection device 100 functions as an acquisition unit 11, a key point detection unit 12, a region segmentation unit 13, a determination unit 14, a color channel conversion unit 15, a binarization processing unit 16, a candidate extraction unit 17, a separation degree calculation unit 18, a color channel selection unit 19, a hand image detection unit 20, a biological information extraction unit 21, a comparison unit 22 and a storage unit 23.

[0026] Figure 1 (b) is a block diagram illustrating the hardware structure of the detection device 100. Figure 1 As illustrated in (b) of FIG. 1 , the detection device 100 includes a CPU 101 , a RAM 102 , a storage device 103 , an interface 104 , a display device 105 , an input device 106 , a camera 107 , and the like.

[0027] CPU (Central Processing Unit) 101 is a central processing unit. CPU 101 includes one or more cores. RAM (Random Access Memory) 102 is a volatile memory that temporarily stores programs executed by CPU 101, data processed by CPU 101, etc. Storage device 103 is a non-volatile storage device. As storage device 103, for example, ROM (Read Only Memory), solid state drives (SSD) such as flash memory, hard disks driven by hard disk drives, etc. can be used. Storage device 103 stores detection programs. Interface 104 is an interface device with external devices. For example, interface 104 is an interface device with LAN (Local Area Network).

[0028] The display device 105 is a display device such as LCD (Liquid Crystal Device), etc. The input device 106 is an input device such as a keyboard and a mouse, etc. The camera 107 is a MOS (Metal Oxide Semiconductor) sensor, a CCD (Charged Coupled Device) sensor, etc.

[0029] The CPU 101 executes the detection program to realize the acquisition unit 11, the key point detection unit 12, the region segmentation unit 13, the determination unit 14, the color channel conversion unit 15, the binarization unit 16, the candidate extraction unit 17, the separation degree calculation unit 18, the color channel selection unit 19, the hand image detection unit 20, the biological information extraction unit 21, the comparison unit 22, and the storage unit 23. In addition, as the acquisition unit 11, the key point detection unit 12, the region segmentation unit 13, the determination unit 14, the color channel conversion unit 15, the binarization unit 16, the candidate extraction unit 17, the separation degree calculation unit 18, the color channel selection unit 19, the hand image detection unit 20, the biological information extraction unit 21, the comparison unit 22, and the storage unit 23, hardware such as dedicated circuits may also be used.

[0030] (Hand image detection processing)

[0031] Figure 2 FIG. 1 is a flowchart showing an example of a hand image detection process performed by the detection device 100. Figure 2 An example of the operation of the detection device 100 is described.

[0032] The acquisition unit 11 acquires an original image including a hand image and a background image from the camera 107 (step S1 ).

[0033] Next, the key point detection unit 12 detects a plurality of key points representing the parts of the hand in the original image acquired by the acquisition unit 11 (step S2). For example, the key point detection unit 12 detects 21 key points P0 to P1. 20 . Key point P0 is a point at the boundary between the palm and the wrist. Key point P1 is a point at the base of the thumb. Key point P2 is a point at the second joint of the thumb. Key point P3 is a point at the first joint of the thumb. Key point P4 is a point at the tip of the thumb. Key point P5 is a point at the base of the index finger. Key point P6 is a point at the second joint of the index finger. Key point P7 is a point at the first joint of the index finger. Key point P8 is a point at the tip of the index finger. Key point P9 is a point at the base of the middle finger. Key point P 10 It is the 1st point of the second joint of the middle finger. Key point P 11 It is the 1st point of the first joint of the middle finger. Key point P 12 It is the point at the tip of the middle finger. Key point P 13 It is the point at the base of the ring finger. Key point P 14 It is the 1 point of the second joint of the ring finger. Key point P 15 It is the 1st point of the first joint of the ring finger. Key point P 16 It is the point at the tip of the ring finger. Key point P 17 It is the point at the base of the little finger. Key point P 18 It is the 1st point of the second joint of the little finger. Key point P19 It is the 1st point of the first joint of the little finger. Key point P 20 It is 1 point of the tip of the little finger.

[0034] These key points P0~P 20 For example, key points P0 to P 20 It is obtained by detecting the bounding box representing the hand area and then detecting each part of the hand. 20 The method of non-patent document 1 "T. Simon, H. Joo, I. Matthews, and Y. Sheikh, "Hand keypoint detection in single images using multiview bootstrapping," in CVPR, 2017" can be used for detection. 20 ,exist Figure 3 exemplified in (a).

[0035] Next, the region segmentation unit 13 obtains a plurality of small regions A to D from the original image using the key points (step S3 ). A small region is a region smaller than the original image. Figure 3 (b) is a diagram illustrating small areas A to D. The black circle drawn by the black line passes through the key point P0, the key point P6, and the key point P 18 The area D is composed of key points P0, P5, and P 17 The small area B is composed of the black circle, the straight line connecting key point P6 and key point P5, and the straight line connecting key point P5 and key point P 13 The straight line connecting the key points P 13 and key point P 14 The small area A is the area on the left side of the black circle compared to the small area B and the small area D. The small area C is the area on the right side of the black circle compared to the small area B and the small area D.

[0036] Next, the determination unit 14 determines whether each small area falls within the hand when viewed from above (step S4). The entire small area D falls within the hand, so the determination is "yes". The small areas A to C include the background, so the determination is "no".

[0037] If the determination result is “YES” in step S4 , the determination unit 14 determines that the small area determined as “YES” in step S4 is a hand area (step S5 ).

[0038] When the result of the judgment in step S4 is "No", the color channel conversion unit 15, the binarization processing unit 16, the candidate extraction unit 17, the separation degree calculation unit 18 and the color channel selection unit 19 separate the small area judged as "No" in step S4 into the hand area and the background area (step S6). The details will be described later.

[0039] Next, the hand image detection unit 20 combines the hand regions of the small regions obtained in step S5 and step S6 to detect a hand image (step S7 ).

[0040] Figure 4 The following is a diagram showing an example of a detailed flowchart of step S6. Figure 4 In this embodiment, the small areas A to C are respectively processed. Figure 4 processing.

[0041] First, the color channel conversion unit 15 converts the small area as the object into multiple color channel images (step S11). The multiple color channel images include a first color channel image and a second color channel image of at least two color channels. Color channels are, for example, R (Red), G (Green), B (Blue), H (Hue), S (Saturation), V (Value·Brightness), Y (Brightness), U (Color difference), V (Color difference), etc. For example, the color channel conversion unit 15 obtains an image of each color channel (hereinafter referred to as a color channel image) for the small area as the object. Among them, the multiple color channel images may also include color channel images of three or more color channels.

[0042] Next, the binarization processing unit 16 performs binarization processing on each color channel image using a predetermined threshold value to obtain a hand candidate and a background candidate (step S12). For example, the hand candidate is represented by white, and the background candidate is represented by black.

[0043] Next, the separation degree calculation unit 18 generates a color histogram (frequency distribution of pixel values) of the hand candidate and a color histogram of the background candidate for each color channel image before the binarization process, and calculates the first separation degree (step S13). For example, the R image is binarized to obtain the hand candidate and the background candidate. Next, the color histogram of the hand candidate and the color histogram of the background candidate are generated for the R image before the binarization process. Based on the overlap of these histograms, the first separation degree of the R image is calculated.

[0044] Next, the separation degree calculation unit 18 performs morphological processing on at least one of the hand candidate and the background candidate for each color channel image after the binarization processing. The so-called morphological processing refers to the filling of the area. The separation degree calculation unit 18 calculates the second separation degree based on the area difference before and after the morphological processing (step S14).

[0045] Next, the color channel selection unit 19 selects the color channel with the maximum sum of the first separation degree and the second separation degree, and obtains the separation result of the palm and the background (step S15). For example, by normalizing the first separation degree and the second separation degree, the first separation degree and the second separation degree can be quantified with the same scale. For example, the first separation degree and the second separation degree are represented by numerical values ​​of 0 to 100, so that the sum of the first separation degree and the second separation degree can be compared with the same reference for each color channel.

[0046] Figure 5 is an example Figure 2 Detailed diagram of steps S5 to S7. Figure 5 As shown in the example, multiple color channel images are obtained for each of the small areas A to C. A color channel image with the maximum sum of the first separation degree and the second separation degree is obtained for each of the small areas A to C. Figure 5 In the example, an R image is selected in small area A, a G image is selected in small area B, and an R image is selected in small area C. Then, the palm area is synthesized based on the binarized image of the selected color channel image. For small area D, the color channel is not selected but the binarization is performed. The synthesized palm area is reshaped in the morphological processing. An area equivalent to the area reshaped in the morphological processing is extracted from the original image, thereby improving the detection accuracy of the hand image. In addition, for example, the hand image is an image representing the palm area.

[0047] Figure 6 to Figure 9 It is a diagram illustrating the details of calculation of the first degree of separation and the second degree of separation. Figure 6 as well as Figure 7 This is a diagram illustrating a case where the degree of separation is large. Figure 8 as well as Fig. 9 This is a diagram illustrating a case where the degree of separation is small.

[0048] Figure 6 is a diagram illustrating a color histogram when the first separation degree is large. Figure 6 In the example, the color histogram of the hand candidate in the color channel image and the color histogram of the background candidate in the color channel image are overlapped. Figure 6As illustrated, the separation of the background candidate and the hand area becomes higher when the binarization threshold is sandwiched. For example, the first separation degree can be defined as the size of the interval between the pixel value at the peak position of the color histogram of the background candidate and the pixel value at the peak position of the color histogram of the hand candidate. The larger the interval, the larger the value of the first separation degree becomes, and the smaller the interval, the smaller the value of the first separation degree becomes. Alternatively, the first separation degree may also be the height of the frequency of the bottom between the two peaks, etc. Alternatively, the first separation degree may also be a value reflecting both the above-mentioned interval and the height of the above-mentioned frequency.

[0049] Figure 7 This is a detailed diagram illustrating a case where the second degree of separation is large. Figure 7 The above figure is a binary image. Figure 7 The following figure is an image obtained by morphological processing of the binary image. If the second separation is large, then Figure 7 As shown in the example, the area difference before and after the morphological processing becomes smaller.

[0050] Figure 8 is a diagram illustrating a color histogram when the first separation degree is small. Figure 8 In the example, the color histogram of the hand candidate in the color channel image and the color histogram of the background candidate in the color channel image are overlapped. If the first separation degree is small, then Figure 8 As shown in the example, in the color histogram, the separation between the background area and the palm area becomes low across the binarization threshold.

[0051] Fig. 9 This is a detailed diagram illustrating a case where the second degree of separation is small. Fig. 9 The above figure is a binary image. Fig. 9 The following figure is an image obtained by morphological processing of the binary image. If the second separation is small, then Fig. 9 As shown in the example, the area difference before and after the morphological processing becomes larger.

[0052] The following describes a biometric authentication process using the detected palm area.

[0053] (Registration Process)

[0054] Fig.10 This is a flowchart showing an example of the registration process executed by the detection device 100.

[0055] like Fig.10 As illustrated, the acquisition unit 11 acquires the original image captured by the camera 107 and the ID (identification information) of the registrant (step S21). The ID is input by the registrant or the like using the input device 106, for example.

[0056] Next, the hand image detection unit 20 detects a hand image through the above-mentioned hand image detection process (step S22).

[0057] Next, the biometric information extraction unit 21 extracts biometric information from the hand image detected in step S22 (step S23). The biometric information is veins, palm prints, etc., but is not particularly limited as long as it is biometric information obtained from the hand image.

[0058] The storage unit 23 associates the biometric information extracted in step S23 with the ID acquired in step S21 and stores it as registered biometric data (step S24). The registration process is completed through the above process. For example, by repeating the registration process multiple times, the biometric information of multiple registrants can be registered.

[0059] (Control treatment)

[0060] Fig.11 This is a flowchart showing an example of the comparison process executed by the detection device 100.

[0061] like Fig.11 As illustrated, the acquisition unit 11 acquires the original image captured by the camera 107 (step S31 ).

[0062] Next, the hand image detection unit 20 detects a hand image through the above-mentioned hand image detection process (step S32).

[0063] Next, the biometric information extraction unit 21 extracts biometric information from the hand image detected in step S32 as biometric data for comparison (step S33 ).

[0064] The comparison unit 22 compares the comparison biometric data extracted in step S33 with each registered biometric data stored in the storage unit 23, thereby performing authentication (step S34). Specifically, it is determined whether the similarity between each registered biometric data stored in the storage unit 23 and the comparison biometric data extracted in step S33 is greater than a threshold value. If there is registered biometric data with a similarity greater than the threshold value, it is determined that the comparison biometric data is the biometric information of the person with the ID of the registered biometric data. The comparison unit 22 outputs the determination result.

[0065] According to the present embodiment, the separation degree, which is an index for determining the separability of the background image and the hand image, is calculated based on the first color channel image and the second color channel image of the original image including the background image and the hand image. Based on the calculated separation degree, a color channel is selected. The hand image included in the original image is detected using the selected color channel. The hand image is detected using a color channel with a high separation degree, so that the hand image and the background image can be accurately separated regardless of the shooting conditions. Therefore, the detection accuracy of the hand image is improved.

[0066] In addition, the degree of separation is not particularly limited as long as it is an index for determining the separability between the background image and the hand image.

[0067] However, by using the first separation degree calculated based on the color histograms calculated for the background candidate and the hand candidate, respectively, the detection accuracy of the separability of the background candidate and the hand candidate is improved. In addition, by using the second separation degree corresponding to the area difference before and after the morphological processing performed on at least one of the background candidate and the hand candidate, the detection accuracy of the separability of the background candidate and the hand candidate is improved.

[0068] In addition, the separability of the background candidates and the hand candidates is determined for each of the multiple small areas obtained from the original image, so that a color channel can be selected for each small area. As a result, the detection accuracy of the hand image from the original image is further improved. Small areas are obtained using key points about the parts of the hand, so that the influence of individual differences in the palm can be reduced. Therefore, the influence of individual differences such as the size and shape of the palm can be reduced. No color channel selection is performed for the small area determined to be located inside the hand, so that the processing amount for detecting the hand image can be reduced.

[0069] In addition, in each of the above examples, the color channel selection unit 19 is an example of a color channel selection unit that selects a color channel based on a separation degree, which is calculated based on the first color channel image and the second color channel image of the original image including the background image and the hand image, and is an indicator for determining the separability of the background image and the hand image. The hand image detection unit 20 is an example of a hand image detection unit that detects the hand image included in the original image using the color channel selected by the color channel selection unit. The binarization processing unit 16 is an example of a binarization processing unit that binarizes the first color channel image and the second color channel image, respectively. The candidate extraction unit 17 is an example of a candidate extraction unit that extracts background candidates and hand candidates from the binarized first color channel image and the second color channel image, respectively. The separation degree calculation unit 18 is an example of a frequency distribution calculation unit that calculates the frequency distribution of pixel values ​​for the background candidates and the hand candidates, respectively. In addition, the separation degree calculation unit 18 is an example of a morphological processing unit that performs morphological processing on at least one of the background candidates and the hand candidates. The region dividing unit 13 is an example of a region dividing unit that divides the first color channel image and the second color channel image into a plurality of regions. The determination unit 14 is an example of a determination unit that determines whether a plurality of regions include backgrounds by using key points.

[0070] As mentioned above, although the embodiment of the present invention is described in detail, the present invention is not limited to the above-mentioned specific embodiment, and various deformation|transformation and changes are possible within the scope of the summary of the present invention described in the claims.

[0071] Description of Reference Numerals

[0072] 11…acquisition unit; 12…key point detection unit; 13…region segmentation unit; 14…determination unit; 15…color channel conversion unit; 16…binarization unit; 17…candidate extraction unit; 18…separation degree calculation unit; 19…color channel selection unit; 20…hand image detection unit; 21…biological information extraction unit; 22…control unit; 23…storage unit; 100…detection device; 105…display device; 107…camera.

Claims

1. A detection device, characterized in that: have: A color channel selection unit selects a color channel based on a separation degree, wherein the separation degree is calculated based on a first color channel image and a second color channel image of an original image including a background image and a hand image, and is an index for determining separability of the background image and the hand image; and The hand image detection unit detects the hand image contained in the original image using the color channel selected by the color channel selection unit.

2. The detection device according to claim 1, characterized in that: have: A binarization processing unit, binarizing the first color channel image and the second color channel image respectively; A candidate extraction unit extracts background candidates and hand candidates from the binarized first color channel image and the second color channel image, respectively; as well as A frequency distribution calculation unit calculates the frequency distribution of pixel values ​​for each of the background candidate and the hand candidate. The separation degrees include a first separation degree calculated according to a frequency distribution of pixel values.

3. The detection device according to claim 2, characterized in that: The first separation degree is a separation degree corresponding to a pixel value difference between two frequency peaks in the frequency distribution of the pixel values.

4. The detection device according to claim 1, characterized in that: have: A binarization processing unit, binarizing the first color channel image and the second color channel image respectively; A candidate extraction unit extracts background candidates and hand candidates from the binarized first color channel image and the second color channel image, respectively; as well as a morphological processing unit that performs morphological processing on at least one of the background candidate and the hand candidate, The separation degree includes a second separation degree corresponding to the area difference before and after the morphological processing.

5. The detection device according to any one of claims 1 to 4, characterized in that: A region segmentation unit is provided, which segments the first color channel image and the second color channel image into a plurality of regions respectively. The color channel selection unit selects a color channel according to the degree of separation of each of the plurality of areas.

6. The detection device according to claim 5, characterized in that: The region segmentation unit segments the first color channel image and the second color channel image into the plurality of regions, respectively, based on key points regarding the hand parts detected from the original image.

7. The detection device according to claim 6, characterized in that: A determination unit is provided, the determination unit determining whether the plurality of regions include a background by using the key point, For the area determined by the determination unit as not including the background, the color channel selection unit does not select a color channel, and the hand image detection unit detects it as the hand image.

8. A detection method, characterized in that: The following processing is performed by a computer: Selecting a color channel based on a separation degree, wherein the separation degree is calculated based on a first color channel image and a second color channel image of an original image including a background image and a hand image, respectively, and is an indicator for determining separability of the background image and the hand image; The hand image contained in the original image is detected using the selected color channel.

9. The detection method according to claim 8, characterized in that: The computer performs the following processing: Binarizing the first color channel image and the second color channel image respectively; extracting a background candidate and a hand candidate from each of the binarized first color channel image and the second color channel image; as well as Calculating the frequency distribution of pixel values ​​for the background candidate and the hand candidate respectively, The separation degrees include a first separation degree calculated according to a frequency distribution of pixel values.

10. The detection method according to claim 9, characterized in that: The first separation degree is a separation degree corresponding to a pixel value difference between two frequency peaks in the frequency distribution of the pixel values.

11. The detection method according to claim 8, characterized in that: The computer performs the following processing: Binarizing the first color channel image and the second color channel image respectively; extracting a background candidate and a hand candidate from each of the binarized first color channel image and the second color channel image; as well as performing morphological processing on at least one of the background candidate and the hand candidate, The separation degree includes a second separation degree corresponding to the area difference before and after the morphological processing.

12. The detection method according to any one of claims 8 to 11, characterized in that: The computer performs the following processing: The first color channel image and the second color channel image are divided into a plurality of regions respectively. A color channel is selected based on the degree of separation of each of the plurality of regions.

13. The detection method according to claim 12, characterized in that: The first color channel image and the second color channel image are respectively divided into the plurality of regions according to key points of the hand parts detected from the original image.

14. The detection method according to claim 13, characterized in that: The computer performs the following processing: By using the key points, determining whether the plurality of regions include backgrounds, For the area determined not to include the background, the color channel is not selected and the area is detected as the hand image.

15. A detection program, characterized in that The computer is made to execute the following processing: Selecting a color channel based on a separation degree, wherein the separation degree is calculated based on a first color channel image and a second color channel image of an original image including a background image and a hand image, respectively, and is an indicator for determining separability of the background image and the hand image; The hand image contained in the original image is detected using the selected color channel.

16. The detection program according to claim 15, characterized in that The computer is caused to execute the following processing: Binarizing the first color channel image and the second color channel image respectively; extracting background candidates and hand candidates from the binarized first color channel image and the second color channel image respectively; as well as Calculating the frequency distribution of pixel values ​​for the background candidate and the hand candidate respectively, The separation degrees include a first separation degree calculated according to a frequency distribution of pixel values.

17. The detection program according to claim 16, characterized in that The first separation degree is a separation degree corresponding to a pixel value difference between two frequency peaks in the frequency distribution of the pixel values.

18. The detection program according to claim 15, characterized in that The computer is caused to execute the following processing: Binarizing the first color channel image and the second color channel image respectively; extracting background candidates and hand candidates from the binarized first color channel image and the second color channel image respectively; as well as performing morphological processing on at least one of the background candidate and the hand candidate, The separation degree includes a second separation degree corresponding to the area difference before and after the morphological processing.

19. The detection program according to any one of claims 15 to 18, characterized in that: The computer is caused to execute the following processing: The first color channel image and the second color channel image are divided into a plurality of regions respectively. A color channel is selected based on the degree of separation of each of the plurality of regions.

20. The detection procedure according to claim 19, characterized in that The first color channel image and the second color channel image are respectively divided into the plurality of regions according to key points of the hand parts detected from the original image.

21. The detection procedure according to claim 20, characterized in that The computer is caused to execute the following processing: Using the key points, determining whether the plurality of regions contain background, For the area determined not to include the background, the color channel is not selected and the area is detected as the hand image.

Images