CN106780559B - Moving target detection method and device - Google Patents
Moving target detection method and device Download PDFInfo
- Publication number
- CN106780559B CN106780559B CN201611237422.6A CN201611237422A CN106780559B CN 106780559 B CN106780559 B CN 106780559B CN 201611237422 A CN201611237422 A CN 201611237422A CN 106780559 B CN106780559 B CN 106780559B
- Authority
- CN
- China
- Prior art keywords
- gray
- optical flow
- formula
- image
- gradient
- 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
Classifications
-
- 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
Abstract
The application provides a moving target detection method and a moving target detection device. The method comprises the following steps: acquiring a video image; carrying out gray level processing on the video image to obtain a gray level image; calculating to obtain a first formula based on the gray level image, a preset CLG optical flow model E and a preset condition; according to the first formula, the optical flow increment du of the k-layer pyramid is solved by adopting a fixed point iteration method according to the linear invariant assumption of gray scalek，dvk(ii) a An optical flow increment du according to the k-layer pyramidk，dvkAnd realizing the detection of the operation target. The influence of illumination change on the detection result is considered, so that the algorithm can evaluate the moving target more accurately, and the moving target can be stably and accurately detected when the illumination change is suddenly changed.
Description
Technical Field
The present application relates to the field of video detection technologies, and in particular, to a method and an apparatus for detecting a moving object.
Background
In the prior art, a CLG (Combined-local-Global) based optical flow model is used to detect a moving object, and in order to reduce the sensitivity of the CLG optical flow model to illumination changes, a gradient conservation assumption is Combined with the CLG optical flow model to obtain a new optical flow model, and an energy equation of the new optical flow model is as follows:
in the above formula, I is a gray scale image,is the optical flow gradient of the gray image in the x and y directions, rho is the standard deviation, KρFor its gaussian weight function, α is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, y (x) is a penalty function,wherein e is 0.0001.
The inventor of the present application finds that the above-mentioned detection of a moving object based on a new optical flow model E is premised on gradient conservation, that is, the detected moving object needs to be guaranteed to be detected in a stable illumination environment, and when the illumination change suddenly changes, for example, the detected object suddenly changes from a dark environment (for example, a room without lighting) to a bright environment (for example, a room with lighting), the gradient conservation is no longer true during the sudden change of illumination, and then the new optical flow model E can no longer detect the moving object.
Disclosure of Invention
In view of this, the present application provides a moving target detection method and apparatus, so as to solve the problem that the prior art cannot detect a moving target when an illumination change changes suddenly. The technical scheme is as follows:
based on one aspect of the present application, the present application provides a moving object detection method, including:
acquiring a video image;
carrying out gray level processing on the video image to obtain a gray level image;
based on the gray level image and a preset CLG optical flow modelAnd preset conditionsCalculating to obtain a first formula; wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively being the gray scale imageOptical flow in the x, y and gray gradient maximum directions, Ixx、Iyy、Ixy、Ixt、IytThe derivatives of the optical flow in the three directions of x, y and t are respectively; the first formula is
According to the first formula, the optical flow increment du of the k-layer pyramid is solved by adopting a fixed point iteration method according to the linear invariant assumption of gray scalek，dvk(ii) a k is a positive integer;
an optical flow increment du according to the k-layer pyramidk，dvkAnd realizing the detection of the operation target.
Preferably, the optical flow increment du of the k-layer pyramid is obtained by adopting a fixed point iteration method according to the first formula and a gray linear invariant assumptionk，dvkThe method comprises the following steps:
obtaining a second formula according to the first formula and a gray linear invariant hypothesis; the second formula isWherein the content of the first and second substances,
and solving the optical flow increment du of the k-layer pyramid by using the second formula and a fixed point iteration methodk，dvk。
Preferably, after performing the gray processing on the video image to obtain a gray image, the method further includes:
judging whether the gray gradient of a pixel point in the gray image is larger than a preset threshold value or not;
the preset CLG optical flow modelThe deformation is as follows:
whereinT is a preset threshold value.
Based on another aspect of the present application, the present application further provides a moving object detecting device, including:
a video image acquisition unit for acquiring a video image;
the gray processing unit is used for carrying out gray processing on the video image to obtain a gray image;
a first calculation unit for calculating a CLG optical flow model based on the gray imageAnd preset conditionsCalculating to obtain a first formula; wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively the optical flow in the x, y and maximum direction of the gray scale gradient, Ixx、Iyy、Ixy、Ixt、IytThe derivatives of the optical flow in the three directions of x, y and t are respectively; the first formula is
A second calculating unit, configured to obtain the optical flow increment du of the k-layer pyramid by using a fixed point iteration method according to the first formula and a gray-scale linear invariant assumptionk，dvk(ii) a k is a positive integer;
an operation target detection unit for increasing du optical flow according to the k-layer pyramidk，dvkAnd realizing the detection of the operation target.
Preferably, the second calculation unit includes:
the first calculation subunit is used for obtaining a second formula according to the first formula and a gray scale linear invariant hypothesis; the second formula isWherein the content of the first and second substances,
a second calculating subunit, configured to use the second formula to obtain the optical flow increment du of the k-layer pyramid by using a fixed-point iteration methodk，dvk。
Preferably, the method further comprises the following steps:
the gray gradient judging unit is used for judging whether the gray gradient of the pixel points in the gray image is greater than a preset threshold value or not;
the preset CLG optical flow modelThe deformation is as follows:
whereinT is a preset threshold value.
By applying the moving target detection method provided by the application, the gray level of the obtained video image is processed to obtain a gray level imagePost-imaging, improved CLG-based optical flow modelAnd realizing the detection of the moving target. The CLG optical flow model provided by the application determines the relative motion of each pixel point in an image by utilizing the correlation of the gray value time sequence change of each pixel point in a video image sequence, projects each pixel point in the video image sequence onto a three-dimensional object, gives a speed vector corresponding to each pixel point and the three-dimensional object, and dynamically analyzes the three-dimensional object according to the characteristics of the speed vector. The influence of illumination change on the detection result is considered, so that the algorithm can evaluate the moving target more accurately, and the moving target can be stably and accurately detected when the illumination change is suddenly changed.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of a moving object detection method provided in the present application;
fig. 2 is a schematic structural diagram of a moving object detection apparatus provided in the present application;
fig. 3 is another schematic structural diagram of a moving object detection apparatus provided in the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 1, a flowchart of a moving object detection method provided in the present application is shown, including:
step 101, acquiring a video image.
And 102, carrying out gray processing on the video image to obtain a gray image.
After each frame of video image is read, the video image can be detected by adopting structural texture decomposition. Structural texture decomposition is introduced into a video sequence preprocessing part, so that the algorithm can better adapt to the change of a dynamic background, and the influence of illumination change on optical flow estimation is effectively reduced.
103, based on the gray level image and a preset CLG optical flow modelAnd preset conditionsAnd calculating to obtain a first formula.
The first formula is
Wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively the optical flow in the x, y and maximum direction of the gray scale gradient, Ixx、Iyy、Ixy、Ixt、IytThe derivatives of the optical flow in the three directions x, y and t, respectively.
Specifically, in the present application, the formula (3) can be obtained by combining the formula (1) and the formula (2).
104, according to the first formula, according to the linear invariant hypothesis of the gray scale, adopting a fixed point iteration method to obtain the optical flow increment du of the k-layer pyramidk，dvk(ii) a k is a positive integer.
Specifically, according to the first formula, the second formula is obtained according to the assumption that the gray scale linearity is unchanged.
The second formula isWherein the content of the first and second substances,
and (4) solving the optical flow increment du of the k-layer pyramid by using the second formula (4) and a fixed point iteration methodk、dvk。
In the present application, the above equation (4) is satisfied only when u and v are sufficiently small in value on the assumption that the gradation linearity is not changed. In order to deal with the situation that the target displacement is large in actual target detection, a fixed point iteration method is adopted to obtain the optical flow increment du of the k-layer pyramidk、dvk。
105, according to the optical flow increment du of the k-layer pyramidk，dvkAnd realizing the detection of the operation target.
Obtaining the light stream increment du of the k-layer pyramidk，dvkAnd then, calculating a background motion vector according to the obtained displacement and amplitude of the optical flow, and completing the detection of the dynamic background motion target.
By applying the moving object detection method provided by the application, after the gray level of the obtained video image is processed to obtain a gray level image, the method is based on the improved CLG optical flow modelAnd realizing the detection of the moving target. CLG optical flow model provided by the application utilizes each image in video image sequenceDetermining the relative motion of each pixel point in the image by the correlation of the time sequence change of the gray value of the pixel point, projecting each pixel point in the video image sequence to the three-dimensional object, giving a corresponding velocity vector of each pixel point and the three-dimensional object, and carrying out dynamic analysis on the three-dimensional object according to the characteristics of the velocity vector. The influence of illumination change on the detection result is considered, so that the algorithm can evaluate the moving target more accurately, and the moving target can be stably and accurately detected when the illumination change is suddenly changed.
In the above embodiment, although the CLG optical flow model provided by the present application can better solve the problem that the optical flow calculation accuracy of the optical flow algorithm is low under the condition of illumination change, in the actual moving object detection process, the algorithm needs to reach a certain number of iterations to meet the accuracy requirement of the required optical flow, which results in a large calculation amount of the algorithm and fails to meet the real-time requirement of the moving object detection. Based on the method, the algorithm is improved in the aspect of calculation amount, so that the real-time performance of the algorithm is improved, and the real-time performance requirement of the system is met.
Specifically, after the video image is subjected to the gray processing to obtain the gray image in step 102, the method may further include step 106, in which whether the gray gradient of the pixel point in the gray image is greater than a preset threshold is determined.
In order to meet the real-time requirement on the detection of a moving object, after a gray image is obtained, firstly, a threshold segmentation method is used for calculating the optical flow of pixel points with the gray gradient larger than a preset threshold T in the gray image by using the improved CLG optical flow algorithm, namely the formula (1), and then, the optical flow of other pixel points is obtained through smooth iteration. Secondly, according to the size of the image, a multilayer Gaussian pyramid structure is adopted for hierarchical calculation, the calculation amount of the algorithm is reduced, and the real-time performance of target detection is improved.
In general, the gray levels of corresponding pixels before and after the target moves remain unchanged, and only the pixel points with larger gray levels (the gray levels are larger than a preset threshold T) are approximately established. Therefore, the CLG optical flow algorithm is only adopted for the pixel points with larger gray gradient, and the optical flows of other pixel points adopt a smooth iteration method, so that the detection precision of the algorithm is ensured.
The present application defines the weight function:
at this time, the above-mentioned preset CLG optical flow modelThe deformation is as follows:
wherein T is a preset threshold.
In this application, an optical flow iteration formula for a pixel point with a small gray scale gradient (the gray scale gradient is less than or equal to a preset threshold T) is as follows:
where k is the number of optical flow iterations, λ is the weighting coefficient,is the local average of the optical flow in both the x and y directions.
Meanwhile, the method of convolution of the Gaussian low-pass filter and the image is adopted, and the original image is subjected to down sampling twice to obtain a 4-layer Gaussian pyramid. For each pixel point in the image, the motion vector g at the bottom layer of the Gaussian pyramid can be made2Obtaining the initial motion vector g of the second layer of Gaussian pyramid through up-sampling1＝2v2. By analogy, the optical flow value of the topmost gaussian pyramid can be obtained:
where L is the pyramid level and d is the motion vector.
For the low-level Gaussian pyramid, the iteration times can be properly increased due to less image pixel points so as to improve the accuracy of the calculated optical flow value. And for the high-level Gaussian pyramid, the iteration times can be reduced to meet the real-time requirement of the algorithm. The results of several consecutive frames in the video image, compared with the above algorithm before and after the improvement of the reduced computation amount, are shown in table 1 below. Table 1 shows the algorithm time consumption comparison before and after the improvement to reduce the amount of computation:
TABLE 1
In the embodiment of the application, after each frame of video image is read, the video image is detected through structural texture decomposition, so that the influence of illumination change on optical flow estimation can be effectively reduced. Secondly, the Gaussian pyramid is adopted to calculate the optical flows in a layered mode, and the optical flows of other pixel points are obtained through optical flow iteration of the pixel points with larger gray gradients, so that the calculation amount of the algorithm is reduced, and the detection efficiency is improved. And finally, calculating a background motion vector according to the obtained displacement and amplitude of the optical flow to finish the detection of the dynamic background motion target.
Based on the moving object detection method provided by the foregoing application, the present application further provides a moving object detection apparatus, as shown in fig. 2, including:
a video image acquisition unit 100 for acquiring a video image;
a gray processing unit 200, configured to perform gray processing on the video image to obtain a gray image;
a first computing unit 300 for generating a CLG optical flow model based on the gray imageAnd preset conditionsCalculating to obtain a first formula; wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively the optical flow in the x, y and maximum direction of the gray scale gradient, Ixx、Iyy、Ixy、Ixt、IytThe derivatives of the optical flow in the three directions of x, y and t are respectively; the first formula is
A second calculating unit 400, configured to, according to the first formula, obtain the optical flow increment du of the k-layer pyramid by using a fixed-point iteration method according to a linear invariant gray scale assumptionk，dvk(ii) a k is a positive integer;
an operation target detection unit 500 for increasing the optical flow du according to the k-level pyramidk，dvkAnd realizing the detection of the operation target.
Wherein the second calculation unit 400 comprises:
a first calculating subunit 401, configured to obtain, according to the first formula, a second formula according to a linear invariant gray scale assumption; the second formula isWherein the content of the first and second substances,
a second calculating subunit 402, configured to use the second formula to obtain the optical flow increment du of the k-layer pyramid by using a fixed-point iteration methodk，dvk。
Preferably, the present application further includes a gray gradient determination unit 600, as shown in fig. 3, where the gray gradient determination unit 600 is specifically configured to determine whether a gray gradient of a pixel point in the gray image is greater than a preset threshold;
at this time, the preset CLG optical flow modelThe deformation is as follows:
whereinT is a preset threshold value.
It should be noted that, in the present specification, the embodiments are all described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other. For the device-like embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The moving object detection method and apparatus provided by the present application are described in detail above, and specific examples are applied in the description to explain the principles and embodiments of the present application, and the description of the above embodiments is only used to help understand the method and core ideas of the present application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.
Claims (6)
1. A moving object detection method, comprising:
acquiring a video image;
carrying out gray level processing on the video image to obtain a gray level image;
based on the gray level image and a preset CLG optical flow modelAnd preset conditionsCalculating to obtain a first formula; wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively the optical flow in the x, y and maximum direction of the gray scale gradient, Ixx、Iyy、Ixy、Ixt、IytThe luminous flux is respectively x, y,Derivatives in the three directions t; the first formula is
According to the first formula, the optical flow increment du of the k-layer pyramid is solved by adopting a fixed point iteration method according to the linear invariant assumption of gray scalek，dvk(ii) a k is a positive integer;
an optical flow increment du according to the k-layer pyramidk，dvkAnd realizing the detection of the operation target.
2. The method according to claim 1, wherein said first formula is used for solving the optical flow increment du of the k-level pyramid by using a fixed-point iteration method according to a gray scale linear invariant assumptionk，dvkThe method comprises the following steps:
obtaining a second formula according to the first formula and a gray linear invariant hypothesis; the second formula isWherein the content of the first and second substances,
and solving the optical flow increment du of the k-layer pyramid by using the second formula and a fixed point iteration methodk，dvk。
3. The method according to claim 1 or 2, wherein after performing the gray processing on the video image to obtain a gray image, the method further comprises:
judging whether the gray gradient of a pixel point in the gray image is larger than a preset threshold value or not;
when the gray gradient of a pixel point in the gray image is larger than a preset threshold value, the preset CLG optical flow modelThe deformation is as follows:
whereinT is a preset threshold value.
4. A moving object detecting apparatus, comprising:
a video image acquisition unit for acquiring a video image;
the gray processing unit is used for carrying out gray processing on the video image to obtain a gray image;
a first calculation unit for calculating a CLG optical flow model based on the gray imageAnd preset conditionsCalculating to obtain a first formula; wherein I is the grayscale image,the optical flow gradient of the gray level image in the x and y directions is shown, rho is a standard deviation, KρIs a Gaussian weight function, alpha is a smoothing term, w ═ u, v,1]TWhere u, v are the velocities of the optical flow in the x, y directions, ψ (x) is a penalty function,epsilon takes a value of 0.0001; i isx、Iy、ItRespectively the optical flow in the x, y and maximum direction of the gray scale gradient, Ixx、Iyy、Ixy、Ixt、IytThe derivatives of the optical flow in the three directions of x, y and t are respectively; the first formula is
A second calculating unit, configured to obtain the optical flow increment du of the k-layer pyramid by using a fixed point iteration method according to the first formula and a gray-scale linear invariant assumptionk，dvk(ii) a k is a positive integer;
an operation target detection unit for increasing du optical flow according to the k-layer pyramidk，dvkAnd realizing the detection of the operation target.
5. The apparatus of claim 4, wherein the second computing unit comprises:
the first calculation subunit is used for obtaining a second formula according to the first formula and a gray scale linear invariant hypothesis; the second formula isWherein the content of the first and second substances,
a second calculating subunit, configured to use the second formula to obtain the optical flow increment du of the k-layer pyramid by using a fixed-point iteration methodk，dvk。
6. The apparatus of claim 4 or 5, further comprising:
the gray gradient judging unit is used for judging whether the gray gradient of the pixel points in the gray image is greater than a preset threshold value or not;
when the gray gradient of a pixel point in the gray image is larger than a preset threshold value, the preset CLG optical flow modelThe deformation is as follows:
whereinT is a preset threshold value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611237422.6A CN106780559B (en) | 2016-12-28 | 2016-12-28 | Moving target detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611237422.6A CN106780559B (en) | 2016-12-28 | 2016-12-28 | Moving target detection method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106780559A CN106780559A (en) | 2017-05-31 |
| CN106780559B true CN106780559B (en) | 2019-12-24 |
Family
ID=58924967
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611237422.6A Active CN106780559B (en) | 2016-12-28 | 2016-12-28 | Moving target detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106780559B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109932281B (en) * | 2017-12-19 | 2021-08-17 | 中国科学院沈阳自动化研究所 | Vision-based liquid viscosity on-line measuring method |
| CN108171725A (en) * | 2017-12-25 | 2018-06-15 | 北京航空航天大学 | Object detection method under a kind of dynamic environment based on normal direction stream information |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102946523A (en) * | 2012-10-31 | 2013-02-27 | 江苏省电力公司信息通信分公司 | Method for de-interlacing unmanned substation surveillance video based on CLG (Combining Local and Global Algorithm) and AVS (Audio Video coding Standard) |
| CN104869387A (en) * | 2015-04-19 | 2015-08-26 | 中国传媒大学 | Method for acquiring binocular image maximum parallax based on optical flow method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2722816A3 (en) * | 2012-10-18 | 2017-04-19 | Thomson Licensing | Spatio-temporal confidence maps |
-
2016
- 2016-12-28 CN CN201611237422.6A patent/CN106780559B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102946523A (en) * | 2012-10-31 | 2013-02-27 | 江苏省电力公司信息通信分公司 | Method for de-interlacing unmanned substation surveillance video based on CLG (Combining Local and Global Algorithm) and AVS (Audio Video coding Standard) |
| CN104869387A (en) * | 2015-04-19 | 2015-08-26 | 中国传媒大学 | Method for acquiring binocular image maximum parallax based on optical flow method |
Non-Patent Citations (2)
| Title |
|---|
| 《Fire detection algorithm based on CLG-TV optical flow model》;Kaixi Li 等;《2016 2nd IEEE International Conference on Computer and Communications》;20161017;全文 * |
| 《改进的CLG变分光流场估计方法》;谭君 等;《电子设计工程》;20160131;第24卷(第1期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106780559A (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111080675B (en) | Target tracking method based on space-time constraint correlation filtering | |
| CN109214380B (en) | License plate inclination correction method | |
| CN106780559B (en) | Moving target detection method and device | |
| CN109003234A (en) | For the fuzzy core calculation method of motion blur image restoration | |
| CN109029425A (en) | A kind of fuzzy star chart restored method filtered using region | |
| Han et al. | An improved corner detection algorithm based on harris | |
| CN112132830A (en) | Air tightness detection water body shaking sensing method and system based on artificial intelligence | |
| Pawaskar et al. | Detection of moving object based on background subtraction | |
| CN108376393A (en) | A kind of blurred picture blind restoration method towards high-speed straight-line Moving Objects | |
| CN101937562A (en) | Construction method for gray-level information content histogram | |
| CN105975911A (en) | Energy perception motion significance target detection algorithm based on filter | |
| CN107958464B (en) | Fast and steady estimation method for optical flow field in noise environment | |
| Liu et al. | Extreme learning machine and moving least square regression based solar panel vision inspection | |
| CN109658357A (en) | A kind of denoising method towards remote sensing satellite image | |
| CN107341818A (en) | Image analysis algorithm for the test of touch-screen response performance | |
| Haifeng et al. | Moving object detection based on background subtraction of block updates | |
| CN110910332B (en) | Visual SLAM system dynamic fuzzy processing method | |
| CN108492308A (en) | A kind of determination method and system of variation light stream based on mutual structure guiding filtering | |
| Gu et al. | Drms: Dim-light robust monocular simultaneous localization and mapping | |
| Zhou et al. | Underwater moving target detection based on image enhancement | |
| Wang | An improved YOLO algorithm for object detection in all day scenarios | |
| Shouhong et al. | A Multi-exposure Image Fusion Method with Detail Preservation | |
| Qiao et al. | On particle filter and mean shift tracking algorithm based on multi-feature fusion | |
| CN109767394B (en) | Blurred image restoration method for non-uniform linear motion | |
| Yuan et al. | A Smoke Recognition Method Combined Dynamic Characteristics and Color Characteristics of Large Displacement Area |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |