CN110708511A - Monitoring video compression method based on image target detection - Google Patents
Monitoring video compression method based on image target detection Download PDFInfo
- Publication number
- CN110708511A CN110708511A CN201910986376.7A CN201910986376A CN110708511A CN 110708511 A CN110708511 A CN 110708511A CN 201910986376 A CN201910986376 A CN 201910986376A CN 110708511 A CN110708511 A CN 110708511A
- Authority
- CN
- China
- Prior art keywords
- video
- frame
- target
- tool
- detection
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
Abstract
The invention discloses a monitoring video compression method based on image target detection, and relates to the technical field of video compression; extracting each frame of image information in the surveillance video, extracting HOG characteristics from each remaining frame of image information, carrying out SVM classifier training and screening by using the HOG characteristics to select pedestrians and/or vehicles as targets for detection, reserving a first frame of image in the surveillance video as a background image, reserving a video frame rate and a video frame number, reserving and recording a frame number and a frame coordinate position of the video frame where the target is detected, intercepting an area where the target is located, deleting the video frame where the target is not detected, and compressing the surveillance video.
Description
Technical Field
The invention discloses a monitoring video compression method based on image target detection, and relates to the technical field of video compression.
Background
Particularly, with the further advance of digital life, a large amount of high-definition monitoring videos are generated in daily life. Compared with the common video, the monitoring video has the following characteristics: firstly, the video background is relatively fixed: because the monitoring camera is often fixed, the background of the monitoring video is basically not changed for a long period of time; secondly, the video content is stable: for the nature of the surveillance video, the content scene of the surveillance video is usually stable, and there is little drastic change, and the number of video frames is usually low. Especially at night, the monitoring video content is not changed; thirdly, the video target is more definite: the monitoring video aim generally has higher pertinence, most of the monitoring video aim focuses on monitoring pedestrians and vehicles, and the video aim is definite. Moreover, the time for monitoring the video is long, the occupied space is large, and the moving or copying needs a long time, so that the video compression is necessary. Video compression is a process of video analysis processing, and a process of changing the format of video content through video coding aims to reduce the storage space occupied by video.
The existing video compression method removes a large amount of spatial and temporal redundant information in a video by the technologies of intra-frame prediction, inter-frame prediction, quantization, coding and the like, but the video compression method takes a frame as a unit and cannot remove content redundant information, so that the method cannot well compress a monitoring video and a large amount of content redundant information still exists in the monitoring video.
The invention determines a monitoring video compression method based on image target detection, which processes the input monitoring video on the basis of utilizing HOG characteristics and an SVM classifier to detect the image target, analyzes and stores the area with the background changed compared with the background of the monitoring video, and deletes redundant information on a large amount of contents in the monitoring video. Compared with the existing video compression method, the method improves the video compression ratio of the monitoring video and improves the storage benefit of the monitoring video.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a monitoring video compression method based on image target detection, which improves the video compression ratio of the monitoring video and improves the storage benefit of the monitoring video compared with the existing video compression method.
The specific scheme provided by the invention is as follows:
a monitoring video compression method based on image target detection extracts the information of each frame of image in the monitoring video,
extracting HOG characteristics from each remaining frame of picture information, training and screening pedestrians and/or vehicles as targets by an SVM classifier by using the HOG characteristics to detect,
reserving a first frame picture in a monitoring video as a background picture, reserving a video frame rate and a video frame number, reserving and recording a frame number of a video frame where a target is detected and a frame coordinate position where the video frame is located, intercepting an area where the target is located,
deleting the video frames without the detected target, and compressing the monitoring video.
In the method, the OpenCV tool in python is used for extracting the picture information of the video frame in the monitoring video and carrying out target detection.
In the method, the video Capture () function and the read () function of the OpenCV tool are used for extracting the picture information of the video frame in the monitoring video.
In the method, the target detection is carried out on the remaining picture information of each frame by utilizing setSVMDetector () and detectMultiScale () in an HOGDescriptor class of an OpenCV tool.
In the method, the region where the target is located is intercepted by utilizing the imwrite () function of python.
In the method, the frame number of the video frame of the target, the coordinate position of the frame, the video frame rate and the video frame number are stored by using a dump () function under a pickle tool of python.
A monitoring video compression tool based on image target detection comprises an extraction module, a detection module and a compression module,
the extraction module extracts the picture information of each frame in the monitoring video,
the detection module extracts HOG characteristics from each frame of image information, and uses the HOG characteristics to train and screen pedestrians and/or vehicles as targets for detection by an SVM classifier,
the compression module reserves a first frame picture in the surveillance video as a background picture, reserves a video frame rate and a video frame number, reserves and records a frame number and a frame coordinate position of a video frame where the target is detected, intercepts an area where the target is located, deletes the video frame where the target is not detected, and compresses the surveillance video.
And the in-tool extraction module extracts the picture information of the video frame in the monitoring video and performs target detection by using an OpenCV tool in python.
The in-tool extraction module extracts picture information of video frames in the monitoring video by using a VideoCapture () function and a read () function of the OpenCV tool.
The detection module in the tool utilizes setSVMDetector () and detectMultiScale () in an HOGDescriptor class of an OpenCV tool to perform target detection on each remaining frame of picture information.
The invention has the advantages that:
the invention provides a monitoring video compression method based on image target detection, which comprises the steps of extracting each frame of image information in a monitoring video, extracting HOG characteristics from each remaining frame of image information, utilizing the HOG characteristics to train an SVM classifier and screen pedestrians and/or vehicles as targets for detection, retaining a first frame of image in the monitoring video as a background image, retaining a video frame rate and a video frame number, simultaneously retaining and recording a frame number of the video frame where the target is detected and a frame coordinate position where the video frame where the target is located, intercepting an area where the target is located, deleting the video frame where the target is not detected, and compressing the monitoring video;
the method processes the input surveillance video on the basis of utilizing the HOG characteristics and the SVM classifier to detect the image target, analyzes and stores the area with the background changed compared with the background of the surveillance video, and deletes redundant information on a large amount of contents in the surveillance video. Compared with the existing video compression method, the method improves the video compression ratio of the monitoring video and improves the storage benefit of the monitoring video.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention.
Detailed Description
The invention provides a monitoring video compression method based on image target detection, which extracts the information of each frame of image in the monitoring video,
extracting HOG characteristics from each remaining frame of picture information, training and screening pedestrians and/or vehicles as targets by an SVM classifier by using the HOG characteristics to detect,
reserving a first frame picture in a monitoring video as a background picture, reserving a video frame rate and a video frame number, reserving and recording a frame number of a video frame where a target is detected and a frame coordinate position where the video frame is located, intercepting an area where the target is located,
deleting the video frames without the detected target, and compressing the monitoring video.
Meanwhile, a monitoring video compression tool based on image target detection corresponding to the method is provided, which comprises an extraction module, a detection module and a compression module,
the extraction module extracts the picture information of each frame in the monitoring video,
the detection module extracts HOG characteristics from each frame of image information, and uses the HOG characteristics to train and screen pedestrians and/or vehicles as targets for detection by an SVM classifier,
the compression module reserves a first frame picture in the surveillance video as a background picture, reserves a video frame rate and a video frame number, reserves and records a frame number and a frame coordinate position of a video frame where the target is detected, intercepts an area where the target is located, deletes the video frame where the target is not detected, and compresses the surveillance video.
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
When the method of the invention is used for compressing the monitoring video, the specific process is as follows:
extracting picture information of video frames in the surveillance video by using a videoCapture () function and a read () function of an OpenCV tool,
and performing target detection on the remaining each frame of picture information by utilizing setSVMDetector () and detectMultiScale () in an HOGDescriptor class of an OpenCV tool: extracting HOG characteristics, training and screening pedestrians and/or vehicles as targets by using an SVM classifier by using the HOG characteristics for detection,
the first frame picture in the surveillance video is reserved as a background picture by utilizing the imwrite () function under the OpenCV tool in python,
intercepting the detected target area, storing the area picture by utilizing an imwrite () function, storing the frame number corresponding to the area picture, the frame position coordinate, the video frame rate and the frame number by utilizing a dump () function under a pickle packet in python,
deleting the video frames without the detected target, and compressing the monitoring video.
When decompression is performed, the background picture information and the target area picture information can be read by using the immed () function under the cv2 packet in python,
reading the frame number, the frame position coordinate, the video frame rate and the frame number information by using a load () function under a pick packet in python,
restoring each frame of video by combining the background picture and the target area picture,
the original surveillance video is restored by combining each frame of picture by using the VideoWriter () function under the cv2 packet in python.
When the tool is used for compressing the monitoring video, the specific process is as follows:
the extraction module extracts picture information of video frames in the surveillance video by using a VideoCapture () function and a read () function of the OpenCV tool,
the detection module performs target detection on each remaining frame of picture information by utilizing setSVMDetector () and detectMultiScale () in an HOGDescriptor class of an OpenCV tool: extracting HOG characteristics, training and screening pedestrians and/or vehicles as targets by using an SVM classifier by using the HOG characteristics for detection,
the compression module retains the first frame picture in the surveillance video as the background picture using the imwrite () function under the OpenCV tool in python,
intercepting the detected target area, storing the area picture by utilizing an imwrite () function, storing the frame number corresponding to the area picture, the frame position coordinate, the video frame rate and the frame number by utilizing a dump () function under a pickle packet in python,
deleting the video frames without the detected target, and compressing the monitoring video.
The tool of the present invention may further comprise a decompression module for decompressing, the decompression module reads the background picture information and the target area picture information by using the immed () function under the cv2 packet in python,
reading the frame number, the frame position coordinate, the video frame rate and the frame number information by using a load () function under a pick packet in python,
restoring each frame of video by combining the background picture and the target area picture,
the original surveillance video is restored by combining each frame of picture by using the VideoWriter () function under the cv2 packet in python.
The above-mentioned embodiments are merely preferred embodiments for fully illustrating the present invention, and the scope of the present invention is not limited thereto. The equivalent substitution or change made by the technical personnel in the technical field on the basis of the invention is all within the protection scope of the invention. The protection scope of the invention is subject to the claims.
Claims (10)
1. A monitoring video compression method based on image target detection is characterized in that the information of each frame of image in the monitoring video is extracted,
extracting HOG characteristics from each remaining frame of picture information, training and screening pedestrians and/or vehicles as targets by an SVM classifier by using the HOG characteristics to detect,
reserving a first frame picture in a monitoring video as a background picture, reserving a video frame rate and a video frame number, reserving and recording a frame number of a video frame where a target is detected and a frame coordinate position where the video frame is located, intercepting an area where the target is located,
deleting the video frames without the detected target, and compressing the monitoring video.
2. The method as claimed in claim 1, wherein the OpenCV tool in python is used to extract the picture information of the video frames in the surveillance video and perform object detection.
3. The method as claimed in claim 2, wherein the video capture () function and the read () function of the OpenCV tool are used to extract the picture information of the video frames in the surveillance video.
4. A method according to claim 2 or 3, characterized by performing object detection on each remaining frame of picture information using setSVMDetector () and detectMultiScale () in the hoddescriptor class of the OpenCV tool.
5. A method as claimed in claim 1 or 4, characterised by intercepting the region of the object using the imwrite () function of python.
6. The method as claimed in claim 5, wherein the frame number of the video frame of the target, the coordinate position of the frame, the video frame rate and the video frame number are stored by using dump () function under a pickle tool of python.
7. A monitoring video compression tool based on image target detection is characterized by comprising an extraction module, a detection module and a compression module,
the extraction module extracts the picture information of each frame in the monitoring video,
the detection module extracts HOG characteristics from each frame of image information, and uses the HOG characteristics to train and screen pedestrians and/or vehicles as targets for detection by an SVM classifier,
the compression module reserves a first frame picture in the surveillance video as a background picture, reserves a video frame rate and a video frame number, reserves and records a frame number and a frame coordinate position of a video frame where the target is detected, intercepts an area where the target is located, deletes the video frame where the target is not detected, and compresses the surveillance video.
8. The tool of claim 7, wherein the extraction module uses an OpenCV tool in python to extract picture information of video frames in the surveillance video and perform object detection.
9. The tool of claim 8, wherein the extraction module extracts the picture information of the video frames in the surveillance video using a VideoCapture () function and a read () function of the OpenCV tool.
10. The tool of claim 8 or 9, wherein the detection module performs the target detection on the remaining picture information of each frame using setSVMDetector () and detectMultiScale () in the hoddescriptor class of the OpenCV tool.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986376.7A CN110708511A (en) | 2019-10-17 | 2019-10-17 | Monitoring video compression method based on image target detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986376.7A CN110708511A (en) | 2019-10-17 | 2019-10-17 | Monitoring video compression method based on image target detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110708511A true CN110708511A (en) | 2020-01-17 |
Family
ID=69201128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910986376.7A Pending CN110708511A (en) | 2019-10-17 | 2019-10-17 | Monitoring video compression method based on image target detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110708511A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111277835A (en) * | 2020-02-18 | 2020-06-12 | 济南浪潮高新科技投资发展有限公司 | Monitoring video compression and decompression method combining yolo3 and flownet2 network |
| CN112954456A (en) * | 2021-03-29 | 2021-06-11 | 深圳康佳电子科技有限公司 | Video data processing method, terminal and computer readable storage medium |
| CN114245070A (en) * | 2021-11-30 | 2022-03-25 | 慧之安信息技术股份有限公司 | Method and system for centralized viewing of regional monitoring content |
| CN114898490A (en) * | 2022-03-31 | 2022-08-12 | 珠海汇金科技股份有限公司 | Security protection method, system, device and storage medium for isolation door |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103475877A (en) * | 2013-09-05 | 2013-12-25 | 广东威创视讯科技股份有限公司 | Video transmission method and system |
| CN103686095A (en) * | 2014-01-02 | 2014-03-26 | 中安消技术有限公司 | Video concentration method and system |
| CN104537688A (en) * | 2014-12-24 | 2015-04-22 | 南京邮电大学 | Moving object detecting method based on background subtraction and HOG features |
| CN104717574A (en) * | 2015-03-17 | 2015-06-17 | 华中科技大学 | Method for fusing events in video summarization and backgrounds |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
| CN108256429A (en) * | 2017-12-19 | 2018-07-06 | 国网山西省电力公司阳泉供电公司 | A kind of transmission tower object detection method using high spatial resolution single polarization SAR image |
| CN108664939A (en) * | 2018-05-16 | 2018-10-16 | 东南大学 | A kind of remote sensing images aircraft recognition method based on HOG features and deep learning |
| CN109246488A (en) * | 2017-07-04 | 2019-01-18 | 北京航天长峰科技工业集团有限公司 | A kind of video abstraction generating method for safety and protection monitoring system |
| CN109498059A (en) * | 2018-12-18 | 2019-03-22 | 首都师范大学 | A kind of contactless humanbody condition monitoring system and body state manage monitoring method |
| CN109615623A (en) * | 2018-12-04 | 2019-04-12 | 广东技术师范学院 | A kind of piping lane video object detection method based on OpenCV |
-
2019
- 2019-10-17 CN CN201910986376.7A patent/CN110708511A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103475877A (en) * | 2013-09-05 | 2013-12-25 | 广东威创视讯科技股份有限公司 | Video transmission method and system |
| CN103686095A (en) * | 2014-01-02 | 2014-03-26 | 中安消技术有限公司 | Video concentration method and system |
| CN104537688A (en) * | 2014-12-24 | 2015-04-22 | 南京邮电大学 | Moving object detecting method based on background subtraction and HOG features |
| CN104717574A (en) * | 2015-03-17 | 2015-06-17 | 华中科技大学 | Method for fusing events in video summarization and backgrounds |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
| CN109246488A (en) * | 2017-07-04 | 2019-01-18 | 北京航天长峰科技工业集团有限公司 | A kind of video abstraction generating method for safety and protection monitoring system |
| CN108256429A (en) * | 2017-12-19 | 2018-07-06 | 国网山西省电力公司阳泉供电公司 | A kind of transmission tower object detection method using high spatial resolution single polarization SAR image |
| CN108664939A (en) * | 2018-05-16 | 2018-10-16 | 东南大学 | A kind of remote sensing images aircraft recognition method based on HOG features and deep learning |
| CN109615623A (en) * | 2018-12-04 | 2019-04-12 | 广东技术师范学院 | A kind of piping lane video object detection method based on OpenCV |
| CN109498059A (en) * | 2018-12-18 | 2019-03-22 | 首都师范大学 | A kind of contactless humanbody condition monitoring system and body state manage monitoring method |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111277835A (en) * | 2020-02-18 | 2020-06-12 | 济南浪潮高新科技投资发展有限公司 | Monitoring video compression and decompression method combining yolo3 and flownet2 network |
| CN112954456A (en) * | 2021-03-29 | 2021-06-11 | 深圳康佳电子科技有限公司 | Video data processing method, terminal and computer readable storage medium |
| CN114245070A (en) * | 2021-11-30 | 2022-03-25 | 慧之安信息技术股份有限公司 | Method and system for centralized viewing of regional monitoring content |
| CN114898490A (en) * | 2022-03-31 | 2022-08-12 | 珠海汇金科技股份有限公司 | Security protection method, system, device and storage medium for isolation door |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110708511A (en) | Monitoring video compression method based on image target detection | |
| US10334249B2 (en) | System and method for high-resolution storage of images | |
| US8953044B2 (en) | Multi-resolution video analysis and key feature preserving video reduction strategy for (real-time) vehicle tracking and speed enforcement systems | |
| Gao et al. | The IEEE 1857 standard: Empowering smart video surveillance systems | |
| CN111901604B (en) | Video compression method, video reconstruction method, corresponding devices, camera and video processing equipment | |
| EP2498223A1 (en) | Video pre-processing before encoding and corresponding post-processing after decoding | |
| US11095899B2 (en) | Image processing apparatus, image processing method, and storage medium | |
| CN103179402A (en) | Video compression coding and decoding method and device | |
| US20150016747A1 (en) | Image processor and image combination method thereof | |
| CN103826109A (en) | Video monitoring image data processing method and system | |
| CN104349074A (en) | Method, device and system for producing a merged digital video sequence | |
| WO2010144176A1 (en) | Method and system for categorized event recording of images in multiple resolution levels | |
| US20200322639A1 (en) | Syntax-based method of detecting object intrusion in compressed video | |
| US20170213086A1 (en) | Cloud platform with multi camera synchronization | |
| JP5503507B2 (en) | Character area detection apparatus and program thereof | |
| CN109583414B (en) | Indoor road occupation detection method, device, medium and processor based on video detection | |
| EP3975133A1 (en) | Processing of images captured by vehicle mounted cameras | |
| US20200221115A1 (en) | Syntax-based Method of Extracting Region of Moving Object in Compressed Video | |
| CN104184997A (en) | Quick surveillance video playback technique | |
| JP2008109247A (en) | Method and device for filtering video noise, integrated circuit, and encoder | |
| CN108282674A (en) | A kind of video transmission method, terminal and system | |
| CN111277835A (en) | Monitoring video compression and decompression method combining yolo3 and flownet2 network | |
| Moriyama et al. | Moving object detection in HEVC video by frame sub-sampling | |
| KR102179077B1 (en) | syntax-based method of providing object classification in compressed video by use of neural network which is learned by cooperation with an external commercial classifier | |
| US20080285955A1 (en) | Method for low data rate video image storage |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200117 |