CN112071006A - High-efficiency low-resolution image area intrusion recognition algorithm and device - Google Patents
High-efficiency low-resolution image area intrusion recognition algorithm and device Download PDFInfo
- Publication number
- CN112071006A CN112071006A CN202010952642.7A CN202010952642A CN112071006A CN 112071006 A CN112071006 A CN 112071006A CN 202010952642 A CN202010952642 A CN 202010952642A CN 112071006 A CN112071006 A CN 112071006A
- Authority
- CN
- China
- Prior art keywords
- human body
- boundary
- image
- body shape
- region
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The invention relates to an efficient low-resolution image area intrusion recognition algorithm and a device, wherein the algorithm specifically comprises the following steps: s1: sampling from a common video stream to obtain two continuous frames of image analysis samples; s2: canny edge detection operation is carried out on the two frames of images, and pixel change areas in the two frames of images are contrasted and analyzed according to Canny edge detection results of the two frames of images; s3: analyzing the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finding out the human body shape position in the human body shape region by using a boundary detection algorithm; s4: and comparing and analyzing the analyzed human body shape and position with the designated area to judge whether the human body invades the designated area. The method can realize the analysis of the low-resolution images, greatly reduces the difficulty and cost of the intrusion identification of the image areas, and has strong applicability and popularization value.
Description
Technical Field
The invention relates to the technical field of computer image processing and recognition, in particular to a high-efficiency low-resolution image area intrusion recognition algorithm and device.
Background
The regional invasion is one of important safety guarantee measures of industrial places and construction sites, the automatic identification of whether a person invades a specified place without permission is a very important means for standardizing and managing the operation behaviors of the construction sites and high-risk places, and the intelligent image analysis and identification are the most direct and effective methods.
The domestic related image area intrusion identification algorithm has been applied for many years, the identification rate is over 95 percent overall, the application effect is good overall, but the algorithm and the system are characterized in that a high-definition camera with higher cost and special hardware are adopted, and the overall cost is higher. The video system resolution of common user units is different, the number of security cameras of social units is large, the video system resolution is specially applied to video recording and preview implementation, the video system resolution is specially applied to the security field, and whether common security video images can be applied to intelligent image analysis, such as image area intrusion identification, work dressing identification and other fields, is a problem which needs to be solved urgently at present.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides a high-efficiency low-resolution image area intrusion identification algorithm and a device, and solves the technical problems that the existing image area intrusion identification algorithm only can adopt a high-definition camera to generate a visual image and the overall cost is high.
The invention is realized by the following technical scheme:
an efficient low-resolution image area intrusion recognition algorithm specifically comprises the following steps:
s1: sampling from a common video stream to obtain two continuous frames of image analysis samples;
s2: canny edge detection operation is carried out on the two frames of images, and pixel change areas in the two frames of images are contrasted and analyzed according to Canny edge detection results of the two frames of images;
s3: analyzing the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finding out the human body shape position in the human body shape region by using a boundary detection algorithm;
s4: and comparing and analyzing the analyzed human body shape and position with the designated area to judge whether the human body invades the designated area.
Further, the image analysis sample in S1 is a 24-bit RGB bitmap, and the image resolution is not lower than 320 × 240.
Further, the step of determining the region range of the boundary of the human body shape in S3 includes:
s31: and (3) carrying out image pixel gray value calculation on the single-frame image sample: AVG _ VALUE ═ R + G + B)/3;
s32: establishing a human body boundary detection model in the image pixel change area, if the human body boundary detection model meets the following three conditions: the boundary smoothness and the boundary trend proportion of the human body boundary detection model are more than 65%, the human body boundary detection model is square or rectangular, and the human body boundary detection model has the characteristics of small top and large bottom and is determined as the region range of the human body shape boundary.
Further, the step of finding the shape position of the human body in S3 includes: after the area range of the human body shape boundary is analyzed, the smoothness of the human body boundary and the trend of the head shoulder boundary are further analyzed, judgment is carried out through the trend of the color pixel points according with the condition, the angle value alpha from the left side to the top of the pixel boundary in the human body shoulder range meets the condition that alpha is more than or equal to 30 degrees and less than or equal to 65 degrees, and the human body is judged.
Further, the specific manner of judging the head shoulder boundary trend is as follows:
judging the human shoulder head from left to right, from bottom to top, from left to right after the human shoulder head reaches the top and from top to bottom;
the judgment of the human shoulder head is from right to left, from bottom to top, from right to left to the top and from top to bottom.
An efficient low-resolution image area intrusion recognition device comprises
The image acquisition module samples and obtains two continuous frames of image analysis samples from a common video stream;
the Canny edge detection module is used for carrying out Canny edge detection operation on the two frames of images and comparing and analyzing pixel change areas in the two frames of images according to Canny edge detection results of the two frames of images;
the human body shape analysis module analyzes the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finds out the human body shape position in the human body shape region by using a boundary detection algorithm;
and the region coincidence comparison analysis module is used for performing coincidence comparison analysis on the analyzed human body shape and position and the designated region and judging whether the human body invades the designated region.
Compared with the prior art, the invention has the beneficial effects that:
according to the high-efficiency low-resolution image area intrusion recognition algorithm and device, a common camera is used for generating a visible video image, the analysis can be performed under the condition of low resolution, the application cost of intelligent image analysis can be effectively reduced, and the popularization rate of the application of the algorithm and device is improved;
the Canny edge detection operation of two continuous frames of images is carried out, the change areas of the two continuous frames of images are compared, the shape of a human body and the positions of the shoulders and the heads in the change areas are analyzed, the contact ratio between the change areas and the defined areas is analyzed to determine whether the areas invade, the algorithm can be applied under the condition of low resolution, and the characteristics of the contact ratio between the shape of the human body and the defined areas are analyzed to automatically identify whether the human body in the images invades the specified areas with high efficiency.
Drawings
Fig. 1 is a schematic flowchart of an efficient low-resolution image area intrusion identification algorithm according to an embodiment of the present invention;
FIG. 2(a) is a diagram of right-to-top angle analysis of pixel boundaries in the shoulder region of a human body according to an embodiment of the present invention;
FIG. 2(b) is a diagram of left-to-top angle analysis of pixel boundaries in the shoulder region of a human body according to an embodiment of the present invention.
Detailed Description
The following examples are presented to illustrate certain embodiments of the invention in particular and should not be construed as limiting the scope of the invention. The present disclosure may be modified from materials, methods, and reaction conditions at the same time, and all such modifications are intended to be within the spirit and scope of the present invention.
As shown in fig. 1, a high-efficiency low-resolution image area intrusion recognition algorithm specifically includes the following steps:
s1: sampling two continuous frames of image analysis samples from a common video stream.
In this embodiment, an image analysis sample can be obtained by a national standard internet video standard protocol or a secondary development kit provided by a video manufacturer, the image analysis sample is a 24-bit RGB bitmap, and the image resolution is not lower than 320 × 240;
s2: canny edge detection operation is carried out on the two frames of images, and pixel change areas in the two frames of images are contrasted and analyzed according to Canny edge detection results of the two frames of images;
in this embodiment, the Canny edge detection operation steps are as follows:
1) establishing a convolution kernel array, and performing Gaussian smoothing on the image to reduce the error rate;
2) performing filtering processing, establishing a gradient direction angle array, and evaluating the edge direction and the intensity of each point pixel by calculating the gradient amplitude and the gradient direction;
3) according to the gradient direction, local non-maximum suppression processing is carried out on the gradient amplitude;
4) and carrying out double-threshold processing, double-threshold middle-threshold filtering, connection and the like on the image edge pixels, and refining the edge pixels by using a Robert operator.
S3: and analyzing the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finding out the human body shape position in the human body shape region by using a boundary detection algorithm.
In this embodiment, the step of determining the region range of the boundary of the human body shape includes:
s31: and (3) carrying out image pixel gray value calculation on the single-frame image sample: AVG _ VALUE ═ R + G + B)/3;
s32: establishing a human body boundary detection model in the image pixel change area, if the human body boundary detection model meets the following three conditions: the boundary smoothness and the boundary trend proportion of the human body boundary detection model are more than 65%, the human body boundary detection model is square or rectangular, and the human body boundary detection model has the characteristics of small top and large bottom and is determined as the region range of the human body shape boundary.
In this embodiment, the step of finding the shape and position of the human body includes: after analyzing the region range of the human body shape boundary, further analyzing the smoothness of the human body boundary and the trend of the head shoulder boundary, and judging by the trend of the color pixel points meeting the conditions, wherein the specific judgment mode is as shown in fig. 2(a) -2 (b):
judging the human shoulder head from left to right, from bottom to top, from left to right after the human shoulder head reaches the top and from top to bottom;
the judgment of the human shoulder head is from right to left, from bottom to top, from right to left to the top and from top to bottom.
The angle value alpha from the left side to the right side to the top of the pixel boundary which accords with the shoulder range of the human body is more than or equal to 30 degrees and less than or equal to 65 degrees, the human body is judged, and if the angle value alpha does not meet the conditions, the human body is not judged.
S4: and comparing and analyzing the analyzed human body shape and position with the designated area, and judging whether the human body invades the designated area or not by analyzing whether the human body shape and position have area overlapping characteristics or not through two continuous frames of images.
If the person invades, starting voice broadcast to remind related personnel of noticing regional invasion, storing image data for later reference, simultaneously sending the image to an on-duty computer and a mobile phone APP of the related personnel, and after receiving image information, carrying out corresponding processing through an internal safety supervision method by a safety supervision personnel, the computer of the on-duty personnel or a mobile terminal; if no person invades, the next analysis is carried out.
An efficient low-resolution image area intrusion recognition device comprises
The image acquisition module samples and obtains two continuous frames of image analysis samples from a common video stream;
the Canny edge detection module is used for carrying out Canny edge detection operation on the two frames of images and comparing and analyzing pixel change areas in the two frames of images according to Canny edge detection results of the two frames of images;
the human body shape analysis module analyzes the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finds out the human body shape position in the human body shape region by using a boundary detection algorithm;
and the region coincidence comparison analysis module is used for performing coincidence comparison analysis on the analyzed human body shape and position and the designated region and judging whether the human body invades the designated region.
In conclusion, the high-efficiency low-resolution image area intrusion identification algorithm and the device have no requirement on the quality of an image, can analyze the image under the condition of low resolution, can achieve the minimum resolution of 320 multiplied by 240 of image pixels, can analyze the shape of a human body and the coincidence degree of an appointed area by using a Canny edge detection algorithm, and can automatically identify whether the human body invades the appointed area with high efficiency and low cost.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by the present specification, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (6)
1. An efficient low-resolution image area intrusion recognition algorithm is characterized by specifically comprising the following steps:
s1: sampling from a common video stream to obtain two continuous frames of image analysis samples;
s2: canny edge detection operation is carried out on the two frames of images, and pixel change areas in the two frames of images are contrasted and analyzed according to Canny edge detection results of the two frames of images;
s3: analyzing the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finding out the human body shape position in the human body shape region by using a boundary detection algorithm;
s4: and comparing and analyzing the analyzed human body shape and position with the designated area to judge whether the human body invades the designated area.
2. The algorithm of claim 1, wherein the image analysis samples in S1 are 24-bit RGB bitmaps, and the image resolution is not lower than 320 × 240.
3. The high-efficiency low-resolution image region intrusion recognition algorithm according to claim 1, wherein the step of determining the region range of the boundary of the human body shape in S3 comprises:
s31: and (3) carrying out image pixel gray value calculation on the single-frame image sample: AVG _ VALUE ═ R + G + B)/3;
s32: establishing a human body boundary detection model in the image pixel change area, if the human body boundary detection model meets the following three conditions: the boundary smoothness and the boundary trend proportion of the human body boundary detection model are more than 65%, the human body boundary detection model is square or rectangular, and the human body boundary detection model has the characteristics of small top and large bottom and is determined as the region range of the human body shape boundary.
4. The high efficiency low resolution image area intrusion recognition algorithm according to any one of claims 1 or 3, wherein the step of finding the shape position of the human body in S3 comprises: after the area range of the human body shape boundary is analyzed, the smoothness of the human body boundary and the trend of the head shoulder boundary are further analyzed, judgment is carried out through the trend of the color pixel points according with the condition, the angle value alpha from the left side to the top of the pixel boundary in the human body shoulder range meets the condition that alpha is more than or equal to 30 degrees and less than or equal to 65 degrees, and the human body is judged.
5. The high-efficiency low-resolution image area intrusion identification algorithm according to claim 4, wherein the specific manner of judging the trend of the head-shoulder boundaries is as follows:
judging the human shoulder head from left to right, from bottom to top, from left to right after the human shoulder head reaches the top and from top to bottom;
the judgment of the human shoulder head is from right to left, from bottom to top, from right to left to the top and from top to bottom.
6. An efficient low-resolution image area intrusion recognition device is characterized by comprising
The image acquisition module samples and obtains two continuous frames of image analysis samples from a common video stream;
the Canny edge detection module is used for carrying out Canny edge detection operation on the two frames of images and comparing and analyzing pixel change areas in the two frames of images according to Canny edge detection results of the two frames of images;
the human body shape analysis module analyzes the region range of the human body shape boundary on the basis of analyzing the change region of the next frame of image, and finds out the human body shape position in the human body shape region by using a boundary detection algorithm;
and the region coincidence comparison analysis module is used for performing coincidence comparison analysis on the analyzed human body shape and position and the designated region and judging whether the human body invades the designated region.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010952642.7A CN112071006A (en) | 2020-09-11 | 2020-09-11 | High-efficiency low-resolution image area intrusion recognition algorithm and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010952642.7A CN112071006A (en) | 2020-09-11 | 2020-09-11 | High-efficiency low-resolution image area intrusion recognition algorithm and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112071006A true CN112071006A (en) | 2020-12-11 |
Family
ID=73696460
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010952642.7A Pending CN112071006A (en) | 2020-09-11 | 2020-09-11 | High-efficiency low-resolution image area intrusion recognition algorithm and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112071006A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105664462A (en) * | 2016-01-07 | 2016-06-15 | 北京邮电大学 | Auxiliary training system based on human body posture estimation algorithm |
US20180047167A1 (en) * | 2016-08-15 | 2018-02-15 | Macau University Of Science And Technology | Method for edge detection |
CN108985144A (en) * | 2018-05-29 | 2018-12-11 | 湖北德强电子科技有限公司 | A kind of high efficiency, low cost image fire automatic identifying method and device |
CN109409190A (en) * | 2018-08-21 | 2019-03-01 | 南京理工大学 | Pedestrian detection method based on histogram of gradients and Canny edge detector |
CN110008831A (en) * | 2019-02-23 | 2019-07-12 | 晋能大土河热电有限公司 | A kind of Intellectualized monitoring emerging system based on computer vision analysis |
CN110598521A (en) * | 2019-07-16 | 2019-12-20 | 南京菲艾特智能科技有限公司 | Behavior and physiological state identification method based on intelligent analysis of face image |
-
2020
- 2020-09-11 CN CN202010952642.7A patent/CN112071006A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105664462A (en) * | 2016-01-07 | 2016-06-15 | 北京邮电大学 | Auxiliary training system based on human body posture estimation algorithm |
US20180047167A1 (en) * | 2016-08-15 | 2018-02-15 | Macau University Of Science And Technology | Method for edge detection |
CN108985144A (en) * | 2018-05-29 | 2018-12-11 | 湖北德强电子科技有限公司 | A kind of high efficiency, low cost image fire automatic identifying method and device |
CN109409190A (en) * | 2018-08-21 | 2019-03-01 | 南京理工大学 | Pedestrian detection method based on histogram of gradients and Canny edge detector |
CN110008831A (en) * | 2019-02-23 | 2019-07-12 | 晋能大土河热电有限公司 | A kind of Intellectualized monitoring emerging system based on computer vision analysis |
CN110598521A (en) * | 2019-07-16 | 2019-12-20 | 南京菲艾特智能科技有限公司 | Behavior and physiological state identification method based on intelligent analysis of face image |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8608073B2 (en) | System and method for robust real-time 1D barcode detection | |
US8532414B2 (en) | Region-of-interest video quality enhancement for object recognition | |
CN101599175B (en) | Detection method for determining alteration of shooting background and image processing device | |
CN110443212B (en) | Positive sample acquisition method, device, equipment and storage medium for target detection | |
CN108389215B (en) | Edge detection method and device, computer storage medium and terminal | |
CN105279487A (en) | Beauty tool screening method and system | |
US11288790B2 (en) | Display panel inspection method and apparatus | |
CN112862832B (en) | Dirt detection method based on concentric circle segmentation positioning | |
CN106228541A (en) | The method and device of screen location in vision-based detection | |
CN113660427A (en) | Image analysis system and method applied to video monitoring tester | |
CN111460988A (en) | Illegal behavior identification method and device | |
Li et al. | A color cast detection algorithm of robust performance | |
CN112862744A (en) | Intelligent detection method for internal defects of capacitor based on ultrasonic image | |
CN110222647B (en) | Face in-vivo detection method based on convolutional neural network | |
CN107292892B (en) | Video frame image segmentation method and device | |
KR20120015980A (en) | Method, system and computer program product for object color correction | |
CN107248151B (en) | Intelligent liquid crystal display detection method and system based on machine vision | |
CN112071006A (en) | High-efficiency low-resolution image area intrusion recognition algorithm and device | |
CN115601712B (en) | Image data processing method and system suitable for site safety measures | |
TW201445458A (en) | Testing device and method for camera | |
CN111241934A (en) | Method and device for acquiring photophobic region in face image | |
CN111708907B (en) | Target person query method, device, equipment and storage medium | |
CN114937003A (en) | Multi-type defect detection system and method for glass panel | |
CN113947563A (en) | Cable process quality dynamic defect detection method based on deep learning | |
CN112329554A (en) | Low-resolution image helmet identification method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201211 |
|
RJ01 | Rejection of invention patent application after publication |