CN105160328B - A kind of human body contour outline recognition methods based on binary image - Google Patents
A kind of human body contour outline recognition methods based on binary image Download PDFInfo
- Publication number
- CN105160328B CN105160328B CN201510592327.7A CN201510592327A CN105160328B CN 105160328 B CN105160328 B CN 105160328B CN 201510592327 A CN201510592327 A CN 201510592327A CN 105160328 B CN105160328 B CN 105160328B
- Authority
- CN
- China
- Prior art keywords
- rectangle frame
- hot spot
- point
- binary image
- human body
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 229910003460 diamond Inorganic materials 0.000 claims abstract description 5
- 239000010432 diamond Substances 0.000 claims abstract description 5
- 238000000605 extraction Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 description 3
- 235000003283 Pachira macrocarpa Nutrition 0.000 description 2
- 241001083492 Trapa Species 0.000 description 2
- 235000014364 Trapa natans Nutrition 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 235000009165 saligot Nutrition 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000005021 gait Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The human body contour outline recognition methods based on binary image that the invention discloses a kind of, this method first obtains video signal from video camera, then binary image is obtained by background difference algorithm, then the rectangle frame for surrounding hot spot is extracted from binary image, by the depth-width ratio for calculating rectangle frame, longitudinal and horizontal line angle α is calculated in rectangle frame of the depth-width ratio more than or equal to 1.64, the ratio that the pixel in diamond accounts for entire rectangle frame pixel is calculated in meeting the rectangle frame that α is 60 ~ 90 degree, the ratio is more than 80% rectangle frame, as human body contour outline；This method has step few, it is easy to operate, the fast advantage of calculating speed, it can quickly and accurately judge which hot spot is personage according to the method for the present invention, interference hot spot and non-subject lighting spot can be removed simultaneously, to rapidly and accurately judge human body contour outline, create conditions to further realize contactless remote identification.
Description
Technical field
The present invention relates to human body contour outline recognition methods, and in particular to a kind of human body contour outline identification side based on binary image
Method.
Background technology
With the rapid development of science and technology, it is widely used in each of society using video camera to monitor dynamic scene
Field, such as relatively high access control system, monitoring system are required to security performance.Since the scene content of video camera shooting is multiple
It is miscellaneous, it needs first to distinguish human body contour outline in clapped scene, it could be further special by the biology such as face picture, fingerprint, iris, gait
Sign further identifies personage, realizes that contactless remote identification, intelligent monitoring, auxiliary are solved a case.
The method of identification manikin has face's method of identification at present.Face recognition method since the shape of face is very unstable,
People can generate many expressions by the variation of face, also very big in the visual pattern difference of different viewing angles, face, and
It is illuminated by the light being affected for condition and facial overcover, it is most difficult to become living things feature recognition field even artificial intelligence field
One of research topic.
Invention content
To solve the above problems, the object of the present invention is to provide a kind of human body contour outline identification side based on binary image
Method.
The present invention to achieve the above object, is achieved through the following technical solutions：
A kind of human body contour outline recognition methods based on binary image, includes the following steps：
1. obtaining video signal from video camera, binary image is then obtained by background difference algorithm；
2. the rectangle frame of hot spot is surrounded in extraction in the binary image of step 1.：By the leftmost point of hot spot and rightmost
Point width W of the distance as hot spot, the height H of the distance of the point of the top and point bottom as hot spot, by hot spot
Vertical line where the point of vertical line, rightmost where leftmost point, the top point where horizontal line where horizontal line and point bottom
Composition surrounds the rectangle frame of hot spot；
3. calculating the depth-width ratio of 2. rectangle frame that step is surrounded outside the hot spot of gained, retain depth-width ratio more than or equal to 1.64
Rectangle frame, for use；
4. being proceeded as follows in rectangle frame of the 3. depth-width ratio that step retains more than or equal to 1.64：Rectangle is found out respectively
The midpoint a of left side hot spot on frame, the midpoint b of the upper right marginal ray spot of rectangle frame, on rectangle frame on the midpoint c and rectangle frame of top hot spot
The midpoint d of following hot spot, connects ab, ac, ad, bc, bd and cd, calculates the angle α of ab and cd, rectangles of the α between 60 ~ 90 degree
Frame is rectangle frame A；
5. being proceeded as follows in the rectangle frame A of step 4.：The water chestnut being made of ac, ad, bd and bc in calculating step 4.
Pixel in shape frame accounts for the ratio of entire rectangle frame pixel, which is more than 80% rectangle frame, as human body contour outline.
Preferably, a kind of human body contour outline recognition methods based on binary image, step 2. in binaryzation in step 1.
The rectangle frame of extraction encirclement hot spot can take following steps in image：
Binary image is scanned from left to right with the straight line of maximum longitudinal length, it, should when scanning to the point intersected with hot spot
Point is the leftmost point of hot spot, records the coordinate of crosspoint, continues to scan to subsequent point, if connected with upper, after
Continuous scanning is until the hot spot that does not connect, by vertical line where the point of vertical line, rightmost where the leftmost point of the hot spot of scanning, most
Horizontal line is connected with horizontal line where point bottom where the point of top, obtains the rectangle frame for surrounding hot spot.
The advantage of the invention is that：
The present invention is to be used in a kind of human body contour outline recognition methods of monolithic machine platform, has step few, easy to operate, meter
Calculate fireballing advantage.It can quickly and accurately judge which hot spot is personage according to the method for the present invention, while interference can be removed
Hot spot and non-subject lighting spot, to rapidly and accurately judge human body contour outline, to further realize contactless remote identity
Identification creates conditions.
Description of the drawings
Fig. 1 is the video signal figure obtained from video camera；
Fig. 2 is binary image；
Fig. 3 is the rectangle frame schematic diagram extracted in binary image；
Fig. 4 is the line schematic diagram at each midpoint in upper and lower, left and right in rectangle frame of the depth-width ratio more than or equal to 1.64；
Fig. 5 is the method schematic diagram for the rectangle frame that hot spot is surrounded in extraction in binary image.
Specific implementation mode
A kind of human body contour outline recognition methods based on binary image, includes the following steps：
1. obtaining video signal from video camera, binary image is then obtained by background difference algorithm；
2. the rectangle frame of hot spot is surrounded in extraction in the binary image of step 1.：By the leftmost point of hot spot and rightmost
Point width W of the distance as hot spot, the height H of the distance of the point of the top and point bottom as hot spot, by hot spot
Vertical line where the point of vertical line, rightmost where leftmost point, the top point where horizontal line where horizontal line and point bottom
Composition surrounds the rectangle frame of hot spot；
3. calculate step 2. gained hot spot outside surround rectangle frame depth-width ratio, removal depth-width ratio be less than 1.64 rectangle
Frame retains the rectangle frame that depth-width ratio is more than or equal to 1.64, for use；
4. being proceeded as follows in rectangle frame of the 3. depth-width ratio that step retains more than or equal to 1.64：Rectangle is found out respectively
The midpoint a of left side hot spot on frame, the midpoint b of the upper right marginal ray spot of rectangle frame, on rectangle frame on the midpoint c and rectangle frame of top hot spot
The midpoint d of following hot spot, connects ab, ac, ad, bc, bd and cd, calculates the angle α of ab and cd, rectangles of the α between 60 ~ 90 degree
Frame is rectangle frame A；
5. being proceeded as follows in the rectangle frame A of step 4.：The water chestnut being made of ac, ad, bd and bc in calculating step 4.
Pixel in shape frame accounts for the ratio of entire rectangle frame pixel, which is more than 80% rectangle frame, as human body contour outline.
Preferably, a kind of human body contour outline recognition methods based on binary image, step 2. in binaryzation in step 1.
The rectangle frame of extraction encirclement hot spot can take following steps in image：
Binary image is scanned from left to right with the straight line of maximum longitudinal length, it, should when scanning to the point intersected with hot spot
Point is the leftmost point of hot spot, records the coordinate of crosspoint, continues to scan to subsequent point, if connected with upper, after
Continuous scanning is until the hot spot that does not connect, by vertical line where the point of vertical line, rightmost where the leftmost point of the hot spot of scanning, most
Horizontal line is connected with horizontal line where point bottom where the point of top, obtains the rectangle frame for surrounding hot spot.
Embodiment 1
A kind of human body contour outline recognition methods based on binary image, includes the following steps：
1. video signal is obtained from video camera, as shown in Figure 1, then converting Fig. 1 to by background difference algorithm
Binary image, binary image as shown in Figure 2；
2. the rectangle frame of hot spot is surrounded in extraction in the binary image of Fig. 2：By the leftmost point of hot spot and rightmost
Height H of the distance of width W of the distance as hot spot of point, the point of the top and point bottom as hot spot, most by hot spot
Vertical line where the point of vertical line, rightmost where the point on the left side, the top point where horizontal line group where horizontal line and point bottom
At the rectangle frame for surrounding hot spot, as shown in Figure 3；
3. calculating the depth-width ratio of the rectangle frame surrounded outside hot spot in Fig. 3, removal depth-width ratio is less than 1.64 rectangle frame, retains
Depth-width ratio is more than or equal to 1.64 rectangle frame, such as there are two rectangle frames in figure 3, and the depth-width ratio of the rectangle frame on the left side is
0.85, give up；The depth-width ratio of the right rectangle frame is 2, is retained for use；
4. being proceeded as follows in rectangle frame of the 3. depth-width ratio that step retains more than or equal to 1.64, as shown in Figure 4：Point
Do not find out the midpoint a of left side hot spot on rectangle frame, the midpoint b of the upper right marginal ray spot of rectangle frame, the midpoint c of top hot spot on rectangle frame
With the midpoint d of marginal ray spot above and below rectangle frame, connect ab, ac, ad, bc, bd and cd, calculate the angle α of ab and cd, retain α 60 ~
The angle of rectangle frame between 90 degree, ab and cd as shown in Figure 4 is 81 degree, is operated into next step；
5. the pixel in the diamond being made of ac, ad, bd and bc in calculating step 4. accounts for entire rectangle frame pixel
Ratio, which is more than 80% rectangle frame, as human body contour outline, as shown in figure 4, the diamond shape being made of ac, ad, bd and bc
Pixel in frame is 896, and total pixel in entire rectangle frame is 985, the diamond being made of ac, ad, bd and bc
In pixel account for entire rectangle frame pixel ratio be 90.9%, then the profile be human body contour outline.
Embodiment 2
The rectangle frame of following methods extraction encirclement hot spot is taken the binary image of Fig. 2 in embodiment 1：As shown in figure 5,
Binary image is scanned from left to right with the straight line of maximum longitudinal length, and when scanning to the point intersected with hot spot, which is hot spot
Leftmost point, record the coordinate of crosspoint, continue to scan to subsequent point, if connected with upper, continue to scan on straight
To the hot spot not connected, by vertical line where the point of vertical line, rightmost where the leftmost point of the hot spot of scanning, the point of the top
Place horizontal line is connected with horizontal line where point bottom, obtains the rectangle frame for surrounding hot spot, the results are shown in Figure 3.
Claims (2)
1. a kind of human body contour outline recognition methods based on binary image, it is characterised in that：Include the following steps：
1. obtaining video signal from video camera, binary image is then obtained by background difference algorithm；
2. the rectangle frame of hot spot is surrounded in extraction in the binary image of step 1.：By the point of hot spot leftmost point and rightmost
Width W of the distance as hot spot, the height H of the distance of the point of the top and point bottom as hot spot, it is most left by hot spot
Vertical line where the point of vertical line, rightmost where the point on side, the top point where horizontal line composition where horizontal line and point bottom
Surround the rectangle frame of hot spot；
3. calculate step 2. gained hot spot outside surround rectangle frame depth-width ratio, retain depth-width ratio be more than or equal to 1.64 rectangle
Frame, for use；
4. being proceeded as follows in rectangle frame of the 3. depth-width ratio that step retains more than or equal to 1.64：It finds out on rectangle frame respectively
The midpoint a of left side hot spot, the midpoint b of the upper right marginal ray spot of rectangle frame, it is following on the midpoint c and rectangle frame of top hot spot on rectangle frame
The midpoint d of hot spot connects ab, ac, ad, bc, bd and cd, calculates the angle α of ab and cd, and rectangle frames of the α between 60 ~ 90 degree is
Rectangle frame A；
5. being proceeded as follows in the rectangle frame A of step 4.：The diamond being made of ac, ad, bd and bc in calculating step 4.
In pixel account for the ratio of entire rectangle frame pixel, which is more than 80% rectangle frame, as human body contour outline.
2. a kind of human body contour outline recognition methods based on binary image according to claim 1, it is characterised in that：Step
2. extraction surrounds the rectangle frame of hot spot and can take following steps in step binary image 1. in：
Binary image is scanned from left to right with the straight line of maximum longitudinal length, and when scanning to the point intersected with hot spot, which is
The leftmost point of hot spot, records the coordinate of crosspoint, continues to scan to subsequent point, if with upper connected, continues to sweep
The hot spot until not connecting is retouched, by vertical line, the top where the point of vertical line, rightmost where the leftmost point of the hot spot of scanning
Point where horizontal line where horizontal line and point bottom connect, obtain the rectangle frame of encirclement hot spot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510592327.7A CN105160328B (en) | 2015-09-17 | 2015-09-17 | A kind of human body contour outline recognition methods based on binary image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510592327.7A CN105160328B (en) | 2015-09-17 | 2015-09-17 | A kind of human body contour outline recognition methods based on binary image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105160328A CN105160328A (en) | 2015-12-16 |
| CN105160328B true CN105160328B (en) | 2018-08-03 |
Family
ID=54801180
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510592327.7A Active CN105160328B (en) | 2015-09-17 | 2015-09-17 | A kind of human body contour outline recognition methods based on binary image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105160328B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358242B (en) * | 2017-07-11 | 2020-09-01 | 浙江宇视科技有限公司 | Target area color identification method and device and monitoring terminal |
| CN108986023A (en) | 2018-08-03 | 2018-12-11 | 北京字节跳动网络技术有限公司 | Method and apparatus for handling image |
| CN109165752A (en) * | 2018-08-21 | 2019-01-08 | 中国平安财产保险股份有限公司 | It is a kind of based on big data car accident processing method and system, communication equipment |
| CN109446895B (en) * | 2018-09-18 | 2022-04-08 | 中国汽车技术研究中心有限公司 | Pedestrian identification method based on human head features |
| CN112507765A (en) * | 2019-09-16 | 2021-03-16 | 瑞昱半导体股份有限公司 | Movement detection method and movement detection system |
| CN116109555B (en) * | 2022-11-16 | 2023-09-29 | 南京博视医疗科技有限公司 | Wavefront facula lattice optimization method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4110736A (en) * | 1974-04-24 | 1978-08-29 | Agency Of Industrial Science & Technology | Shape recognition system |
| CN102799857A (en) * | 2012-06-19 | 2012-11-28 | 东南大学 | Video multi-vehicle outline detection method |
| CN103176668A (en) * | 2013-03-07 | 2013-06-26 | 广东威创视讯科技股份有限公司 | Shot image correction method for camera locating touch system |
| CN104239865A (en) * | 2014-09-16 | 2014-12-24 | 宁波熵联信息技术有限公司 | Pedestrian detecting and tracking method based on multi-stage detection |
-
2015
- 2015-09-17 CN CN201510592327.7A patent/CN105160328B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4110736A (en) * | 1974-04-24 | 1978-08-29 | Agency Of Industrial Science & Technology | Shape recognition system |
| CN102799857A (en) * | 2012-06-19 | 2012-11-28 | 东南大学 | Video multi-vehicle outline detection method |
| CN103176668A (en) * | 2013-03-07 | 2013-06-26 | 广东威创视讯科技股份有限公司 | Shot image correction method for camera locating touch system |
| CN104239865A (en) * | 2014-09-16 | 2014-12-24 | 宁波熵联信息技术有限公司 | Pedestrian detecting and tracking method based on multi-stage detection |
Non-Patent Citations (1)
| Title |
|---|
| 一种动态梯度算法图像二值化边缘提取;曹建秋等;《计算机以信息技术》;20090920;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105160328A (en) | 2015-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105160328B (en) | A kind of human body contour outline recognition methods based on binary image | |
| CN108764058B (en) | Double-camera face in-vivo detection method based on thermal imaging effect | |
| CN106548182B (en) | Pavement crack detection method and device based on deep learning and main cause analysis | |
| KR101355974B1 (en) | Method and devices for tracking multiple object | |
| CN101739548B (en) | Eye detection method and system | |
| CN108268850B (en) | Big data processing method based on image | |
| CN105046206B (en) | Based on the pedestrian detection method and device for moving prior information in video | |
| CN105469113A (en) | Human body bone point tracking method and system in two-dimensional video stream | |
| JPH11288459A (en) | Method and device for detecting area such as face and observer tracking display | |
| CN103971135A (en) | Human body target detection method based on head and shoulder depth information features | |
| KR20090084563A (en) | Method and apparatus for generating the depth map of video image | |
| KR20100072772A (en) | Method and apparatus for real-time face detection using stereo vision | |
| JP2006171929A (en) | Facial area estimation system, facial area estimation method and facial area estimation program | |
| CN103218605A (en) | Quick eye locating method based on integral projection and edge detection | |
| CN105893944A (en) | Depth information static gesture segmentation method | |
| CN102663760A (en) | Location and segmentation method for windshield area of vehicle in images | |
| CN110544300B (en) | Method for automatically generating three-dimensional model based on two-dimensional hand-drawn image characteristics | |
| CN110009650B (en) | Escalator handrail boundary area border crossing detection method and system | |
| CN101650782A (en) | Method for extracting front human face outline based on complexion model and shape constraining | |
| CN105374051B (en) | The anti-camera lens shake video moving object detection method of intelligent mobile terminal | |
| CN110991398A (en) | Gait recognition method and system based on improved gait energy map | |
| CN102073872A (en) | Image-based method for identifying shape of parasite egg | |
| KR20100056143A (en) | Method for detecting moving objects using segmentation process | |
| CN113673614B (en) | Metro tunnel foreign matter intrusion detection device and method based on machine vision | |
| CN105023280A (en) | Detection method based on 3D skeleton modeling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |