WO2005093656A1 - 特定部分姿勢推定装置、特定部分姿勢推定方法及び特定部分姿勢推定プログラム - Google Patents
特定部分姿勢推定装置、特定部分姿勢推定方法及び特定部分姿勢推定プログラム Download PDFInfo
- Publication number
- WO2005093656A1 WO2005093656A1 PCT/JP2004/004040 JP2004004040W WO2005093656A1 WO 2005093656 A1 WO2005093656 A1 WO 2005093656A1 JP 2004004040 W JP2004004040 W JP 2004004040W WO 2005093656 A1 WO2005093656 A1 WO 2005093656A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- specific
- matching
- value
- luminance
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
Definitions
- the present invention is capable of easily and easily estimating a posture in a short time by using an image obtained by a software having a low recognition ability such as a mobile phone or an electronic organizer, or by a simple image processing circuit provided in such a device. It concerns a possible posture estimation device. Background art
- the face of a person including its posture is extracted, image-processed, and used for monitoring or authentication. ing. In such monitoring or authentication, it is necessary to detect the posture first, or it is important to detect the posture before the subsequent image processing.
- an image processing apparatus 21 includes a skin color extracting means 22 for extracting a skin color from an input image, a binarizing means 23 for extracting a candidate region of an eye or a mouth from the result, and a binarizing method. It consists of an eye / mouth area detection / tracking means 24 for detecting eye and mouth areas from the results of the means and tracking them from a moving image.
- the flesh color extraction means 22 detects flesh color pixels
- the binarization means 23 binarizes the flesh color pixels and other pixels.
- the eye / mouth area detection / tracking means 24 extracts a hole area in the skin color area and sets it as a candidate area for eyes and mouth. From the extracted candidate regions, for example, the positions of the eyes and the mouth are detected based on heuristics of the position of the eyes and the position of the mouth with respect to the face region. Finally, head posture information is extracted from the detected eye and mouth positions.
- the conventional posture extraction device is configured as described above, and extracts the skin color of the image of the object, detects and tracks the area, and finally detects the posture.
- the shooting location is changed, it is not constant including the lighting, and it is not always possible to obtain a good image quality, and it is difficult to detect a good skin color.
- the processing amount is large, including the area detection, and a large-scale circuit is required, there is a problem that it takes time to process small-scale hardware installed in a mobile phone.
- the present invention has been made to solve the above-described problems, and an attitude can be extracted in a short time without being affected by the image quality of an input image obtained by a video camera or the like, and by using small-scale hardware having low computational power. I do.
- the specific portion posture estimating apparatus binarizes each pixel of the input image for which the specific portion is estimated based on a set threshold value based on an image value within a predetermined coordinate range, and further, the size is within the set range.
- a pattern 'matching unit which detects a posture by comparing the binary image obtained by the identification with a predetermined template. Further, the matching image generation unit converts the input image into a grayscale image, and calculates the luminance of the grayscale image as an average value or a median value of the luminance of a certain pixel within a predetermined range with the target pixel as a coordinate center. The threshold value is used as a threshold value for binarization. Further, the pattern matching unit generates a matching image from a plurality of specific input images in advance, and uses the generated binary image as a template element. Also, the pattern matching unit is characterized in that the matching with each element of the template is obtained by a logical product of pixels.
- the pattern matching unit obtains pixels having specific values from the binary image obtained by the matching image generation unit, and calculates the inclination of the specific portion to be detected from the state of distribution of those pixels in the image. It is characterized by being estimated.
- the specific portion posture estimating method includes: binarizing each pixel of an input image from which a specific portion is extracted with a set threshold based on an image value within a predetermined coordinate range;
- the input image is converted to a grayscale image, and the luminance of the converted grayscale image is calculated by calculating the average luminance value of pixels within a predetermined range with the target pixel as the coordinate center, and the image median value.
- One of the two is selected as a binarization threshold, and is binarized.
- the collation is characterized in that a collation value is obtained by a luminance logical AND of each pixel corresponding to a coordinate value between the binary image and the template image.
- the specific-part posture estimation program according to the present invention is executable by a computer, and is capable of executing each of the pixels of the input image from which the specific part is extracted by using a set threshold based on an image value within a predetermined coordinate range.
- the program is characterized by comprising a binarized image composed of a group of parts labeled after the above-mentioned deletion, and collating the luminance of the image in a predetermined template.
- FIG. 1 is a diagram showing a configuration of a specific portion posture estimating apparatus according to Embodiment 1 of the present invention.
- FIG. 2 is a flowchart showing an operation of the specific-part-orientation estimation device according to the first embodiment.
- FIG. 3 is a flowchart showing a binarizing operation performed by the matching image generating unit according to the first embodiment.
- FIG. 4 is a diagram illustrating a hardware internal configuration of the binary image generation unit according to the first embodiment.
- FIG. 5 is a diagram illustrating a range in which a set threshold value is obtained in the first embodiment.
- FIG. 6 is a diagram illustrating how to obtain a set threshold value according to the first embodiment.
- FIG. 7 is a diagram for explaining a binarizing operation performed by the matching image generating unit according to the first embodiment.
- FIG. 8 is a flowchart of a matching operation performed by the matching unit according to the first embodiment.
- FIG. 9 is a diagram for explaining a matching operation performed by the pattern matching unit according to the first embodiment.
- FIG. 10 is a diagram showing a configuration of another specific portion posture estimation device according to the first embodiment.
- FIG. 11 is a diagram for explaining a template creation operation performed by another pattern / matching unit according to the first embodiment.
- FIG. 12 is a diagram illustrating a configuration of a specific-part-orientation estimation device according to the second embodiment.
- FIG. 13 is a diagram for describing posture extraction by a pixel distribution performed by a pattern ′ matching unit according to the second embodiment.
- FIG. 14 is a diagram showing a configuration for posture extraction by a conventional image processing apparatus.
- FIG. 1 is a diagram showing a configuration of a specific portion posture estimating apparatus according to the present embodiment of the present invention.
- a specific portion posture estimating apparatus 1 is a video capture unit 2 for capturing a video signal captured by a video camera or the like, performs filtering processing on the captured video, and an image for collating with a posture pattern described later.
- a color space conversion unit 5 for converting the color image captured by the video capture unit 2 to a grayscale image
- a binary image generation unit 6 for converting the converted grayscale image to a binary image
- a region is obtained by integrating adjacent pixels from the image binarized by the binary image generation unit 6, and the part candidate extraction unit 7 that extracts only candidate regions that can be eyes and mouths and the matching image generation unit 3 generate the regions.
- a matching unit 8 for matching a matching image with a previously stored posture pattern image, and a matching pattern DB 9 for storing a pattern used by the matching unit 8.
- FIG. 2 is a flowchart for explaining the operation.
- FIG. 3 is a flowchart for explaining the operation of the binary image generation unit 6.
- FIG. 4 is a diagram showing the internal configuration of the hardware of the binary image generation unit 6, although other elements are the same.
- FIGS. 5 and 6 show the flow of processing performed by the matching image generation unit 3.
- FIG. 7 is a diagram for explaining the matching process performed by the pattern matching unit 4.
- FIG. 8 is a flowchart showing the operation of the matching process performed by the pattern matching unit 4.
- FIG. 9 is a diagram for explaining the operation of the pattern matching unit 4.
- the binary image generator 6 includes a processor 61 and a memory 6
- the processor 61 first reads the grayscale image of the capture image obtained via the input / output interface 64 into the memory 62. Then, the read grayscale image is binarized in accordance with the luminance in S1-3 in FIG. 2 as described later by the steps written in the binarization program 63.
- the video signal is captured by the video capture unit 2 (step S1-1), and the captured color image is converted into a grayscale image by the color space conversion unit 5 (step S1-2).
- G (x, y) is the luminance value at the coordinate value (x, y)
- R, G, B (x, y) is the pixel value of the color image at the coordinate value (x, y).
- a coefficient value when converting a color image to a grayscale image a value other than the above values may be used.
- the color / grayscale conversion in the color space conversion unit 5 may be normalized using the following (Equation 2) and then converted using the above (Equation 1).
- r (x, y) R (x, y) / (R (x, y) IG (x, y) + B (x, y))
- the binary image generating means 6 binarizes the gray scale image in accordance with the luminance (step S1-3).
- a binary image is generated according to the input image.
- the threshold value used as the threshold for the binarization is set to the window coordinates 31 in a predetermined range shown in Fig. 5, and the average luminance value of all pixels within the coordinate range, 25 pixels in Fig. 5, or Find the median and use this value as the threshold.
- the processing shown in FIG. 6 is performed, and the luminance of the target pixel 32 is compared with a threshold value.
- the input image is scanned, and the image is binarized by repeatedly performing the processing from step S2_1 to step S2-8 for all pixels.
- C is a prescribed value set in advance.
- step S2-6 If the condition of (Equation 3) is satisfied, the pixel value is set to 0 (step S2-6); otherwise, the pixel value is set to 1 (step S2-7).
- binarization processing is performed according to the state of the surrounding pixels, for example, a predetermined coordinate range is narrowed even for an image having a low contrast due to deterioration of the image quality of the video camera, so that the range is adaptively adjusted.
- binarization processing can be performed from the average value.
- the average value was obtained in step S2-4.
- the median when the elementary values are rearranged may be obtained, and the following condition (Equation 4) may be used.
- binarization may be performed using a fixed threshold.
- the component candidate extraction unit 7 determines whether the same binary pixel is connected to the binary image 11 in four or eight directions vertically, horizontally, and diagonally, and determines the related and significant adjacent pixels.
- the regions are integrated to obtain the regions, which are labeled as individual regions, such as 1 l_a, 11-b in FIG. 7 (S 1-4). Further, only the area where the size of the circumscribed rectangle of the area falls within the preset range in each area 111a is extracted (step S1-5). That is, the region l l — a in FIG. 7 is excluded because it is a region having a size outside the estimation target.
- the user when considering the use of the videophone function in a mobile phone or an electronic organizer, the user needs to know in advance how large the eyes or mouth should be in order to capture his / her own face within the angle of view and take a large image. Can be predicted. Therefore, the above-described threshold processing is effective.
- the result extracted at step S 1-5 is as shown at 12 in FIG. 7.
- the pattern matching unit 4 estimates the head posture (step S 1-6 ).
- this pattern matching unit 4 is also the configuration shown in FIG. Same as the above, but there is a matching program that performs the operation in FIG. 8 instead of the binarization program 63.
- the template shown in 14 in FIG. 9 is stored in the matching pattern DB9.
- the angle of view can be assumed in advance, so the state of the eyes and mouth area according to the direction of the face Can be predicted in advance.
- the matching pattern DB9 a binary mask image of the eye and mouth regions in the assumed head posture is stored.
- step S3-1 the binarized matching image 13 P in FIG. 9 is read into the memory via the input / output interface.
- the logical product of the matching image 13 and each mask image of the template is calculated, the number of pixels 1 (matched images) of the resulting image is calculated and added, and the image with the largest number is selected.
- the matching result 15 can be obtained.
- the binarization makes the detection of pattern matching not an analog comparison but can be performed very easily.
- a predetermined pattern is stored in the matching pattern DB.
- a template image of the user may be generated by using an image acquired first from the video camera.
- FIG. 10 is a configuration diagram for extracting a specific partial posture when a matching pattern is created from the output of the video capture unit 2.
- a matching pattern generator 16 for generating a template image of a posture pattern from a captured image is provided.
- FIG. 11 is a diagram showing a result of generating a template image based on the image obtained by binarizing an image captured in the normal posture by the video capture unit 2.
- the image captured first in the video capture unit 2 is regarded as a normal posture (a posture facing the camera front), or the user is requested to take a photograph in a normal posture, and an image in the normal posture is acquired. I do.
- the image 17 obtained as described above is binarized by the matching image generation unit 3 using the affine transformation in the matching template generation unit 16 using, for example, an image in which the head is swung right and left, and the head is swung right and left. Generate an image that has been shaken up and down.
- the affine transformation can be represented by the matrix shown in the following equation.
- the user may take various postures as shown in 18 of FIG. 11 and binarize the postures. Then you can generate templates without affine transformation.
- the head pose is estimated by comparing the matching pattern with the matching image.
- the matching means is designed to estimate the head pose from the distribution of pixels having a pixel value of 1 in the matching image. An example in which is changed will be described.
- FIG. 12 is a diagram showing a configuration of the specific portion posture estimation device according to the present embodiment.
- a pixel distribution measuring unit 19 is provided to obtain the pixel distribution of the matching image and estimate the head posture based on the state of the distribution.
- FIG. 13 shows a map for estimating the head posture according to the pixel distribution.
- the processing can be further simplified, and therefore, the processing can be further shortened even with hardware having low computational power.
- the specific partial posture estimating apparatus has been described as being constituted by hardware, but as shown in FIG. 4, a program may be actually prepared and the processor may execute the program. Alternatively, a method including steps representing the flows of FIGS. 2, 3, and 8 may be employed. Industrial applicability
- a matching image generation unit that specifies a component by binarizing an input image based on an average image within a predetermined range, a matching image generation unit that obtains the obtained binary image and a predetermined template Since a pattern matching unit is provided to detect the posture by comparing the postures, it is possible to easily estimate the partial posture while suppressing the scale.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Geometry (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2004800281870A CN1860501A (zh) | 2004-03-24 | 2004-03-24 | 特定部分姿势推测装置、特定部分姿势推测方法及特定部分姿势推测程序 |
PCT/JP2004/004040 WO2005093656A1 (ja) | 2004-03-24 | 2004-03-24 | 特定部分姿勢推定装置、特定部分姿勢推定方法及び特定部分姿勢推定プログラム |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2004/004040 WO2005093656A1 (ja) | 2004-03-24 | 2004-03-24 | 特定部分姿勢推定装置、特定部分姿勢推定方法及び特定部分姿勢推定プログラム |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2005093656A1 true WO2005093656A1 (ja) | 2005-10-06 |
Family
ID=35056398
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2004/004040 WO2005093656A1 (ja) | 2004-03-24 | 2004-03-24 | 特定部分姿勢推定装置、特定部分姿勢推定方法及び特定部分姿勢推定プログラム |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN1860501A (ja) |
WO (1) | WO2005093656A1 (ja) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011128990A (ja) * | 2009-12-18 | 2011-06-30 | Canon Inc | 画像処理装置とその方法 |
KR101298024B1 (ko) * | 2010-09-17 | 2013-08-26 | 엘지디스플레이 주식회사 | 사용자 동적 기관 제스처 인식 방법 및 인터페이스와, 이를 사용하는 전기 사용 장치 |
KR101298023B1 (ko) * | 2010-09-17 | 2013-08-26 | 엘지디스플레이 주식회사 | 사용자 동적 기관 제스처 인식 방법 및 인터페이스와, 이를 사용하는 전기 사용 장치 |
CN106033544B (zh) * | 2015-03-18 | 2020-03-24 | 成都理想境界科技有限公司 | 基于模板匹配的试卷内容区域提取方法 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS55146573A (en) * | 1979-05-03 | 1980-11-14 | Sumitomo Electric Ind Ltd | Binary circuit |
JPH07181012A (ja) * | 1993-12-22 | 1995-07-18 | Nissan Motor Co Ltd | 画像データの特徴量検出装置 |
JPH0981756A (ja) * | 1995-09-14 | 1997-03-28 | Mitsubishi Electric Corp | 顔画像処理装置 |
JPH10143661A (ja) * | 1996-11-11 | 1998-05-29 | Matsushita Electric Ind Co Ltd | データ処理装置 |
JPH11265452A (ja) * | 1998-03-17 | 1999-09-28 | Toshiba Corp | 物体認識装置および物体認識方法 |
JP2001216518A (ja) * | 2000-02-07 | 2001-08-10 | Fuji Photo Film Co Ltd | マッチング方法および装置並びに記録媒体 |
-
2004
- 2004-03-24 CN CNA2004800281870A patent/CN1860501A/zh active Pending
- 2004-03-24 WO PCT/JP2004/004040 patent/WO2005093656A1/ja active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS55146573A (en) * | 1979-05-03 | 1980-11-14 | Sumitomo Electric Ind Ltd | Binary circuit |
JPH07181012A (ja) * | 1993-12-22 | 1995-07-18 | Nissan Motor Co Ltd | 画像データの特徴量検出装置 |
JPH0981756A (ja) * | 1995-09-14 | 1997-03-28 | Mitsubishi Electric Corp | 顔画像処理装置 |
JPH10143661A (ja) * | 1996-11-11 | 1998-05-29 | Matsushita Electric Ind Co Ltd | データ処理装置 |
JPH11265452A (ja) * | 1998-03-17 | 1999-09-28 | Toshiba Corp | 物体認識装置および物体認識方法 |
JP2001216518A (ja) * | 2000-02-07 | 2001-08-10 | Fuji Photo Film Co Ltd | マッチング方法および装置並びに記録媒体 |
Also Published As
Publication number | Publication date |
---|---|
CN1860501A (zh) | 2006-11-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105740780B (zh) | 人脸活体检测的方法和装置 | |
JP5629803B2 (ja) | 画像処理装置、撮像装置、画像処理方法 | |
JP6125188B2 (ja) | 映像処理方法及び装置 | |
JP5445460B2 (ja) | なりすまし検知システム、なりすまし検知方法及びなりすまし検知プログラム | |
JP4860749B2 (ja) | 画像中の人物における位置決め指示との適合性を判定する機器、システム、及び方法 | |
JP5366756B2 (ja) | 情報処理装置及び情報処理方法 | |
US20160162673A1 (en) | Technologies for learning body part geometry for use in biometric authentication | |
JP5725012B2 (ja) | 異物判定装置、異物判定方法および異物判定プログラム | |
JP2005056387A (ja) | 画像処理装置、撮像装置、画像処理方法 | |
JP4682820B2 (ja) | オブジェクト追跡装置及びオブジェクト追跡方法、並びにプログラム | |
JP5170094B2 (ja) | なりすまし検知システム、なりすまし検知方法およびなりすまし検知用プログラム | |
JP2003244521A (ja) | 情報処理方法、装置および記録媒体 | |
JP2004094491A (ja) | 顔向き推定装置および顔向き推定方法ならびに顔向き推定プログラム | |
US7460705B2 (en) | Head-top detecting method, head-top detecting system and a head-top detecting program for a human face | |
CN107944395B (zh) | 一种基于神经网络验证人证合一的方法及系统 | |
JP2007052609A (ja) | 手領域検出装置及び手領域検出方法、並びにプログラム | |
CN110909561A (zh) | 眼睛状态检测系统及眼睛状态检测系统的操作方法 | |
JP4729188B2 (ja) | 視線検出装置 | |
CN111639582B (zh) | 活体检测方法及设备 | |
JP2017027492A (ja) | 顔照合装置およびこれを備えた顔照合システムならびに顔照合方法 | |
JP4691570B2 (ja) | 画像処理装置およびオブジェクト推定プログラム | |
JP2008035301A (ja) | 移動体追跡装置 | |
Panetta et al. | Unrolling post-mortem 3D fingerprints using mosaicking pressure simulation technique | |
JP4390487B2 (ja) | 頭部領域抽出装置 | |
JP4141090B2 (ja) | 画像認識装置、陰影除去装置、陰影除去方法及び記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200480028187.0 Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): BW GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 1020067017711 Country of ref document: KR |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: DE |
|
122 | Ep: pct application non-entry in european phase |