CN112509000A - Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video - Google Patents
Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video Download PDFInfo
- Publication number
- CN112509000A CN112509000A CN202011313802.XA CN202011313802A CN112509000A CN 112509000 A CN112509000 A CN 112509000A CN 202011313802 A CN202011313802 A CN 202011313802A CN 112509000 A CN112509000 A CN 112509000A
- Authority
- CN
- China
- Prior art keywords
- quasi
- real
- frame
- image
- target object
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a moving target tracking algorithm of a multi-path 4K quasi-real-time spliced video, belonging to the field of image processing; relates to a target tracking algorithm technology under a dynamic background; the method is used for solving the problems of low target tracking accuracy and easy loss in the prior art under the dynamic background, and simultaneously provides an instructive detection and tracking method for the tracking of any target under the dynamic and static background.
Description
Technical Field
The invention belongs to the field of image processing; relates to a target tracking algorithm technology under a dynamic background; in particular to a moving target tracking algorithm of a multi-path 4K quasi-real-time spliced video.
Background
Today's intellectuality, along with the vigorous development of unmanned aerial vehicle technique, the technique for security protection and target tracking is changing day by day. Automated target tracking technologies are receiving increasing attention and attention. In general, target tracking is based on japanese-mark tracking in a static background, and a region of change can be obtained by subtracting two images in front and back of a video, and a target can be captured from optical flow characteristics of the region. The main disadvantage of this method is that the effect in a dynamic background is often poor based on a static background; the existing algorithm can be used for capturing a moving target object under a dynamic background in scenes with small camera shake and slow movement, and the number of pixels occupied by the target object is required to be as large as possible. However, when the object is shot at high altitude, the number of pixels occupied by the object is often small, and if the object is enlarged, distortion occurs, which is not favorable for tracking and feature extraction. The target tracking under the dynamic background in the actual scene is more general and has more practical application value, and the effect of continuous target object tracking in criminal investigation and case detection based on unmanned aerial vehicle shooting in the dynamic scene is more and more prominent.
Disclosure of Invention
The invention aims to provide a moving target tracking algorithm of a multi-path 4K quasi-real-time spliced video, which is used for solving the problems of low target tracking accuracy and easy loss in the prior art under a dynamic background.
The purpose of the invention can be realized by the following technical scheme:
the moving target tracking algorithm of the multi-path 4K quasi-real-time spliced video comprises the following steps:
the method comprises the following steps: establishing data connection with a video processing module, and sending the 4K quasi-real-time spliced video data to the video processing module;
step two: the video processing module completes detection and tracking of a tracking target in the 4K quasi-real-time spliced video through a moving target tracking model;
step three: the detection tracking information is sent to the monitor.
Further, the target tracking model comprises an initial layer, a cutting layer, an expansion layer, a correction layer and a searching layer;
the initial layer is used for reading 4K quasi-real-time spliced video data, acquiring a first frame image in the 4K quasi-real-time spliced video data, determining initial position information of a target object to be tracked in the first frame image, selecting a second frame image if the target object to be tracked is not located in the first frame image, and repeating the steps until the target object to be tracked appears, wherein the initial position information comprises coordinates of an upper right corner point in an image coordinate system, the length of the target object to be tracked and the width of the target object to be tracked, and the whole target object is required to be ensured to be located in a rectangular frame;
the cutting layer is used for cutting a rectangular area with the length of L and the width of W in the first frame original image according to the position, the length and the width of the target object to be tracked, and the corresponding unit is the number of pixel points; the rectangular area cut out by the first frame is a standard detection area, correction is not needed, and correction is carried out on each frame after the second frame;
the expansion layer is used for correcting pixels corresponding to the target object as the foreground for the first time to expand the connected region, the expanded seed points are the middle points of the rectangular region, the pixels of the background are not expanded, and the geometric center of the expanded connected region is found by adopting a clustering idea;
the correction layer is used for correcting the position of the target object for the second time, namely realizing correction under the condition of first correction, and the adopted method is that the center of a new rectangular area is determined by a center positioning method of a straight line associated structure, and the coordinate relation of the correction layer in the global image is solved; based on the relation of the new rectangular area coordinates, cutting out a new rectangular frame from the original image to finish the second correction;
and the searching layer compares the similarity in the image of the next frame based on the corrected image, so that the target object of the next frame can be found, and the frame selection is carried out until the detection and the tracking are finished.
Furthermore, the standard rectangular area is the geometric center of the target object to be captured, which is located at the center of the rectangular frame after clipping, that is, the standard rectangular area
And
where column denotes the number of columns of the rectangular area and row denotes the number of rows of the rectangular area.
Further, the seed point is specifically a seed point selected from a central point of the rectangular frame, and 8 neighborhood expansion is performed.
Furthermore, the straight line related structure is an element composed of straight lines formed by pixel compositions in the connected region, and the longest straight line is found in four directions of 15 degrees, 60 degrees, 105 degrees and 135 degrees.
Further, the mode of finding the longest straight line is a mode of traversing round training comparison, finding the straight line segment which is longest in the four directions, taking the center of the straight line segment as the center of the new rectangular area, and solving the coordinate relation of the straight line segment in the global image.
Furthermore, the central point of the connected region is found in the expanded connected region by adopting a clustering method, namely the mean value of the horizontal coordinates and the mean value of the vertical coordinates of all the connected points are solved,
wherein the mean of the abscissa is:
mean ordinate values are:
determining the center of a new cutting area based on the point po (xo, yo), wherein column is the number of columns of the image corresponding to the rectangular frame, and row is the number of rows of the image corresponding to the rectangular frame; and n is the number of the middle foreground pixel points in the connected region.
Further, the x value of the coordinate of the upper left corner of the first updated clipping rectangular area in the original image is:
y has a value of
Further, the length of the straight line in the direction of 15 degrees for the second correction is:
the length of the straight line in the 60-degree direction is:
the length of the line in the 105 degree direction is:
the length of the straight line in the 135-degree direction is as follows:
wherein columns is expressed as the column number of the image, rows is expressed as the row number of the image, and g (i, j) represents the gray pixel value corresponding to the pixel point (i, j).
10. The moving object tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 1, wherein the x value of the original image coordinate at the upper left corner in the second updating clipping rectangular area is:
the value of y is:
compared with the prior art, the invention has the beneficial effects that:
the invention solves the problems of low target tracking accuracy and easy loss in the prior art under a dynamic background, and provides an instructive detection and tracking method for tracking any target under a dynamic and static background.
Drawings
In order to facilitate understanding for those skilled in the art, the present invention will be further described with reference to the accompanying drawings.
Fig. 1 is a schematic block diagram of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings of the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Thus, the detailed description of the embodiments of the present invention provided in the following drawings is not intended to limit the scope of the invention as claimed, but is merely representative of selected embodiments of the invention.
As shown in fig. 1, the moving target tracking algorithm for a multi-path 4K quasi-real-time spliced video includes:
the method comprises the following steps: establishing data connection with a video processing module, and sending the 4K quasi-real-time spliced video data to the video processing module;
step two: the video processing module completes detection and tracking of a tracking target in the 4K quasi-real-time spliced video through a moving target tracking model;
step three: the detection tracking information is sent to the monitor.
The target tracking model comprises an initial layer, a cutting layer, an expansion layer, a correction layer and a searching layer;
the initial layer is used for reading 4K quasi-real-time spliced video data, acquiring a first frame image in the 4K quasi-real-time spliced video data, determining initial position information of a target object to be tracked in the first frame image, selecting a second frame image if the target object to be tracked is not located in the first frame image, and repeating the steps until the target object to be tracked appears, wherein the initial position information comprises coordinates of an upper right corner point in an image coordinate system, the length of the target object to be tracked and the width of the target object to be tracked, and the whole target object is required to be ensured to be located in a rectangular frame;
the cutting layer is used for cutting a rectangular area with the length of L and the width of W in the first frame original image according to the position, the length and the width of the target object to be tracked, and the corresponding unit is the number of pixel points; the rectangular area cut out by the first frame is a standard detection area, correction is not needed, and correction is carried out on each frame after the second frame;
the expansion layer is used for correcting pixels corresponding to the target object as the foreground for the first time to expand the connected region, the expanded seed points are the middle points of the rectangular region, the pixels of the background are not expanded, and the geometric center of the expanded connected region is found by adopting a clustering idea;
the correction layer is used for correcting the position of the target object for the second time, namely realizing correction under the condition of first correction, and the adopted method is that the center of a new rectangular area is determined by a center positioning method of a straight line associated structure, and the coordinate relation of the correction layer in the global image is solved; based on the relation of the new rectangular area coordinates, cutting out a new rectangular frame from the original image to finish the second correction;
and the searching layer compares the similarity in the image of the next frame based on the corrected image, so that the target object of the next frame can be found, and the frame selection is carried out until the detection and the tracking are finished.
The standard rectangular area is the geometric center of the target object to be captured, which is located at the center of the rectangular frame after cutting, namely
And
where column denotes the number of columns of the rectangular area and row denotes the number of rows of the rectangular area.
The seed point is specifically that the central point of the rectangular frame is selected as the seed point, and 8 neighborhood expansion is carried out.
The straight line related structure is an element which is composed of straight lines formed by pixel compositions in a connected region, and the longest straight line is found in four directions of 15 degrees, 60 degrees, 105 degrees and 135 degrees.
The mode of finding the longest straight line is a mode of traversing round training comparison, finding the longest straight line segment in the four directions, taking the center of the straight line segment as the center of a new rectangular area, and solving the coordinate relation of the straight line segment in the global image.
Finding the central point of the connected region by clustering method in the expanded connected region, i.e. solving the mean value of the abscissa and the mean value of the ordinate of all the connected points,
wherein the mean of the abscissa is:
mean ordinate values are:
determining the center of a new cutting area based on the point po (xo, yo), wherein column is the number of columns of the image corresponding to the rectangular frame, and row is the number of rows of the image corresponding to the rectangular frame; and n is the number of the middle foreground pixel points in the connected region.
Updating the x value of the coordinate of the upper left corner of the cutting rectangular area in the original image for the first time, wherein the x value is as follows:
y has a value of
The length of the straight line in the direction of 15 degrees for the second correction is as follows:
the length of the straight line in the 60-degree direction is:
the length of the line in the 105 degree direction is:
the length of the straight line in the 135-degree direction is as follows:
wherein columns is expressed as the column number of the image, rows is expressed as the row number of the image, and g (i, j) represents the gray pixel value corresponding to the pixel point (i, j).
And updating the x value of the upper left corner in the clipping rectangular area in the original image coordinate for the second time as follows:
the value of y is:
the invention is implemented as follows: firstly, establishing data connection with a video processing module, and sending 4K quasi-real-time spliced video data to the video processing module; the video processing module completes detection and tracking of a tracking target in the 4K quasi-real-time spliced video through a moving target tracking model, and finally, detection and tracking information is sent to a monitor.
The above formulas are all calculated by taking the numerical value of the dimension, the formula is a formula which obtains the latest real situation by acquiring a large amount of data and performing software simulation, and the preset parameters in the formula are set by the technical personnel in the field according to the actual situation.
In the embodiments provided by the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and there may be other divisions when the actual implementation is performed; the modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the method of the embodiment.
It will also be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.
The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above examples are only intended to illustrate the technical process of the present invention and not to limit the same, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical process of the present invention without departing from the spirit and scope of the technical process of the present invention.
Claims (10)
1. The moving target tracking algorithm of the multi-path 4K quasi-real-time spliced video is characterized by comprising the following steps:
the method comprises the following steps: establishing data connection with a video processing module, and sending the 4K quasi-real-time spliced video data to the video processing module;
step two: the video processing module completes detection and tracking of a tracking target in the 4K quasi-real-time spliced video through a moving target tracking model;
step three: the detection tracking information is sent to the monitor.
2. The moving object tracking algorithm of the multi-channel 4K quasi-real-time spliced video according to claim 1, wherein the object tracking model comprises an initial layer, a clipping layer, an extension layer, a correction layer and a search layer;
the initial layer is used for reading 4K quasi-real-time spliced video data, acquiring a first frame image in the 4K quasi-real-time spliced video data, determining initial position information of a target object to be tracked in the first frame image, selecting a second frame image if the target object to be tracked is not located in the first frame image, and repeating the steps until the target object to be tracked appears, wherein the initial position information comprises coordinates of an upper right corner point in an image coordinate system, the length of the target object to be tracked and the width of the target object to be tracked, and the whole target object is required to be ensured to be located in a rectangular frame;
the cutting layer is used for cutting a rectangular area with the length of L and the width of W in the first frame original image according to the position, the length and the width of the target object to be tracked, and the corresponding unit is the number of pixel points; the rectangular area cut out by the first frame is a standard detection area, correction is not needed, and correction is carried out on each frame after the second frame;
the expansion layer is used for correcting pixels corresponding to the target object as the foreground for the first time to expand the connected region, the expanded seed points are the middle points of the rectangular region, the pixels of the background are not expanded, and the geometric center of the expanded connected region is found by adopting a clustering idea;
the correction layer is used for correcting the position of the target object for the second time, namely realizing correction under the condition of first correction, and the adopted method is that the center of a new rectangular area is determined by a center positioning method of a straight line associated structure, and the coordinate relation of the correction layer in the global image is solved; based on the relation of the new rectangular area coordinates, cutting out a new rectangular frame from the original image to finish the second correction;
and the searching layer compares the similarity in the image of the next frame based on the corrected image, so that the target object of the next frame can be found, and the frame selection is carried out until the detection and the tracking are finished.
3. The moving target tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 2, wherein the standard rectangular area is a geometric center of a target object to be captured, and is located at the center of the rectangular frame after being cut, namely the center
And
where column denotes the number of columns of the rectangular area and row denotes the number of rows of the rectangular area.
4. The moving object tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 2, wherein the seed point is specifically selected as a center point of a rectangular frame, and 8-neighborhood expansion is performed.
5. The moving object tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 2, wherein the straight-line related structure is an element consisting of straight lines formed by pixel compositions in a connected region, and the longest straight line is found in four directions of 15 degrees, 60 degrees, 105 degrees and 135 degrees.
6. The multi-channel 4K quasi real-time stitched video moving-target tracking algorithm of claim 5, wherein the finding the longest straight line is performed by performing a traversal round training comparison, finding the longest straight line segment in the four directions, and finding the coordinate relationship of the new rectangular area by using the center of the straight line segment as the center of the new rectangular area.
7. The moving object tracking algorithm for multi-channel 4K quasi-real-time spliced video according to claim 2, wherein the center point of the connected region is found in the expanded connected region by clustering, i.e. the mean value of the abscissa and the mean value of the ordinate of all the connected points are obtained,
wherein the mean of the abscissa is:
mean ordinate values are:
determining the center of a new cutting area based on the point po (xo, yo), wherein column is the number of columns of the image corresponding to the rectangular frame, and row is the number of rows of the image corresponding to the rectangular frame; and n is the number of the middle foreground pixel points in the connected region.
8. The moving object tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 1, wherein the x value of the coordinate of the upper left corner of the first updated clipping rectangular area in the original image is:
y has a value of
9. The moving target tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 8, wherein the length of the straight line in the direction of the second correction 15 degrees is:
the length of the straight line in the 60-degree direction is:
the length of the line in the 105 degree direction is:
the length of the straight line in the 135-degree direction is as follows:
wherein columns is expressed as the column number of the image, rows is expressed as the row number of the image, and g (i, j) represents the gray pixel value corresponding to the pixel point (i, j).
10. The moving object tracking algorithm for the multi-channel 4K quasi-real-time spliced video according to claim 1, wherein the x value of the original image coordinate at the upper left corner in the second updating clipping rectangular area is:
the value of y is:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011313802.XA CN112509000A (en) | 2020-11-20 | 2020-11-20 | Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011313802.XA CN112509000A (en) | 2020-11-20 | 2020-11-20 | Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112509000A true CN112509000A (en) | 2021-03-16 |
Family
ID=74959226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011313802.XA Pending CN112509000A (en) | 2020-11-20 | 2020-11-20 | Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112509000A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105282496A (en) * | 2014-12-02 | 2016-01-27 | 四川浩特通信有限公司 | Method for tracking target video object |
| CN109711320A (en) * | 2018-12-24 | 2019-05-03 | 兴唐通信科技有限公司 | A kind of operator on duty's unlawful practice detection method and system |
| CN110415269A (en) * | 2019-07-19 | 2019-11-05 | 浙江大学 | A kind of target tracking algorism under dynamic static background |
-
2020
- 2020-11-20 CN CN202011313802.XA patent/CN112509000A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105282496A (en) * | 2014-12-02 | 2016-01-27 | 四川浩特通信有限公司 | Method for tracking target video object |
| CN109711320A (en) * | 2018-12-24 | 2019-05-03 | 兴唐通信科技有限公司 | A kind of operator on duty's unlawful practice detection method and system |
| CN110415269A (en) * | 2019-07-19 | 2019-11-05 | 浙江大学 | A kind of target tracking algorism under dynamic static background |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101923845B1 (en) | Image processing method and apparatus | |
| CN110390640B (en) | Template-based Poisson fusion image splicing method, system, equipment and medium | |
| US20220036589A1 (en) | Multispectral camera external parameter self-calibration algorithm based on edge features | |
| CN114565863B (en) | Real-time generation method, device, medium and equipment for orthophoto of unmanned aerial vehicle image | |
| CN110415269A (en) | A kind of target tracking algorism under dynamic static background | |
| CN104182968A (en) | Method for segmenting fuzzy moving targets by wide-baseline multi-array optical detection system | |
| CN113850136A (en) | Yolov5 and BCNN-based vehicle orientation identification method and system | |
| CN113873259A (en) | Panoramic video frame insertion method and device and computer readable storage medium | |
| CN114418839A (en) | Image stitching method, electronic device and computer-readable storage medium | |
| CN104104911B (en) | Timestamp in panoramic picture generating process is eliminated and remapping method and system | |
| CN111758118B (en) | Visual positioning method, device, equipment and readable storage medium | |
| CN116433822B (en) | Neural radiation field training method, device, equipment and medium | |
| CN112215205A (en) | Target identification method and device, computer equipment and storage medium | |
| CN112509000A (en) | Moving target tracking algorithm for multi-path 4K quasi-real-time spliced video | |
| CN112669346B (en) | Pavement emergency determination method and device | |
| CN115984124A (en) | Method and device for de-noising and super-resolution of neuromorphic pulse signals | |
| CN116128919A (en) | Multi-temporal image abnormal target detection method and system based on polar constraint | |
| CN111988520B (en) | Picture switching method and device, electronic equipment and storage medium | |
| CN117011481A (en) | Method and device for constructing three-dimensional map, electronic equipment and storage medium | |
| CN104182993B (en) | Target tracking method | |
| CN112488076A (en) | Face image acquisition method, system and equipment | |
| CN118379315B (en) | 8-Direction Sobel edge detection system based on FPGA | |
| CN111010558B (en) | Stumpage depth map generation method based on short video image | |
| CN116883627B (en) | Unmanned aerial vehicle video augmented reality processing method and system | |
| CN113592929B (en) | Unmanned aerial vehicle aerial image real-time splicing method and system |
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: 20210316 |