CN113487660A - Depth information fused moving target detection method, device, medium and equipment - Google Patents
Depth information fused moving target detection method, device, medium and equipment Download PDFInfo
- Publication number
- CN113487660A CN113487660A CN202110670125.5A CN202110670125A CN113487660A CN 113487660 A CN113487660 A CN 113487660A CN 202110670125 A CN202110670125 A CN 202110670125A CN 113487660 A CN113487660 A CN 113487660A
- Authority
- CN
- China
- Prior art keywords
- image
- depth information
- pixel
- moving target
- detected
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 89
- 238000012545 processing Methods 0.000 claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 26
- 230000014759 maintenance of location Effects 0.000 claims abstract description 17
- 239000000203 mixture Substances 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims abstract description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 18
- 238000012937 correction Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000013135 deep learning Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 230000035945 sensitivity Effects 0.000 abstract description 10
- 230000006870 function Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 6
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
Abstract
The invention discloses a moving target detection method, a moving target detection device, a computer readable storage medium and terminal equipment which are integrated with depth information, wherein the method comprises the following steps: acquiring an image to be detected; carrying out depth information extraction processing on the image to be detected to obtain depth information of N pixel points; wherein N > 0; correcting parameters in preset M mixed Gaussian models by using depth information of N pixel points to obtain M moving target detection models; wherein M > 0; performing high contrast retention processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one; obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models; the embodiment of the invention can adjust the sensitivity of the Gaussian mixture model according to the depth information adaptability, and improve the detection precision of the moving target area, thereby reducing the false detection rate and the false alarm rate.
Description
Technical Field
The invention relates to the technical field of moving object detection, in particular to a moving object detection method and device integrating depth information, a computer readable storage medium and terminal equipment.
Background
The moving target detection function in the intelligent video monitoring system is a basic function, and the function plays a supporting role in subsequent functions of detecting target classification, tracking, identification and the like, so that the moving target can be accurately detected, and the subsequent processing of the video monitoring system is facilitated. However, the detection of the moving target is easily interfered by dynamic backgrounds such as illumination, leaf shaking and the like, so that the false detection rate and the false alarm rate of the moving target are increased.
In the prior art, sensitivity is mainly set according to the area size of a moving target area, and the false detection rate and the false alarm rate of moving target detection are reduced through the sensitivity, however, the area of the moving target area set by the method is constant, and when the area of the moving target area is too small, the false detection rate of the moving target is increased; when the area of the moving target area is set to be too large, the missing rate of the moving target is increased.
Disclosure of Invention
The technical problem to be solved in the embodiments of the present invention is to provide a moving target detection method and apparatus, a computer-readable storage medium, and a terminal device, which are capable of adaptively adjusting the sensitivity of a gaussian mixture model according to depth information, and improving the detection accuracy of a moving target region, thereby reducing the false detection rate and the false alarm rate of a moving target.
In order to solve the above technical problem, an embodiment of the present invention provides a moving object detection method fusing depth information, including:
acquiring an image to be detected;
carrying out depth information extraction processing on the image to be detected to obtain depth information of N pixel points; wherein N > 0;
correcting parameters in preset M mixed Gaussian models by using depth information of N pixel points to obtain M moving target detection models; wherein M > 0;
performing high contrast retention processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models.
Further, the depth information includes N pixel depth values, and each pixel point corresponds to one pixel depth value; then, the depth information extraction processing is performed on the image to be detected, so as to obtain the depth information of N pixel points, and the method specifically includes:
and extracting pixel depth values corresponding to N pixel points in the image to be detected through a body sensor, a deep learning algorithm or a binocular vision technology to obtain depth information of the N pixel points.
Further, the correcting the parameters in the preset M mixed gaussian models by using the depth information of the N pixel points to obtain M moving object detection models specifically includes:
correcting a variance threshold and a noise variance threshold in a Gaussian mixture model corresponding to each pixel point according to the pixel depth value corresponding to each pixel point to obtain M moving target detection models;
the variance threshold value in the moving target detection model corresponding to the nth pixel point is the variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; the noise variance threshold value in the moving target detection model corresponding to the nth pixel point is the noise variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; n is more than or equal to N and more than 0.
Further, the high contrast retaining processing is performed on the image to be detected to obtain N high-frequency characteristic values, and the method specifically includes:
filtering the image to be detected by using a preset filter to obtain a blurred image; the method comprises the following steps that pixel points in an image to be detected correspond to pixel points in a blurred image one by one;
and obtaining N high-frequency characteristic values according to the image to be detected, the blurred image and a preset high-contrast retention algorithm.
Further, the obtaining N high-frequency characteristic values according to the image to be detected, the blurred image and a preset high contrast retention algorithm specifically includes:
calculating high-frequency characteristic values of N pixel points in the image to be detected according to a calculation formula of a high-contrast retention algorithm to obtain N high-frequency characteristic values; wherein the high contrast preserving algorithm is calculated as follows:
G(n)=T1(n)-T(n)+A;
g (n) is the nth high-frequency characteristic value; t1(n) is the pixel value corresponding to the nth pixel point in the blurred image; t (n) is a pixel value corresponding to the nth pixel point in the image to be detected; a is a constant greater than 0.
Further, after obtaining the moving object region according to the N high-frequency feature values and the M moving object detection models, the method further includes:
and performing area statistic correction on each pixel point in the moving target area to obtain a corrected moving target area.
Further, the method carries out regional statistical correction on any pixel point in the moving target region through the following steps:
acquiring an m × m neighborhood of the pixel point; wherein, the m-m neighborhood takes the pixel point as a central pixel point; m is an integer greater than 0;
counting the number num0 of background pixels contained in the m × m neighborhood;
when m x m a is not more than num0, correcting the central pixel point as a background pixel point, otherwise correcting the central pixel point as a moving pixel point; wherein a is more than 0 and less than 1.
In order to solve the above technical problem, an embodiment of the present invention further provides a moving object detection apparatus fusing depth information, including:
the image acquisition module is used for acquiring an image to be detected;
the depth information extraction module is used for extracting depth information of the image to be detected to obtain depth information of N pixel points; wherein N > 0;
the model correction module is used for correcting parameters in M preset mixed Gaussian models by using the depth information of the N pixel points to obtain M moving target detection models; wherein M > 0;
the image processing module is used for performing high contrast retention processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and the image detection module is used for obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models.
An embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a stored computer program; when the computer program runs, the computer program controls the device where the computer readable storage medium is located to execute any one of the above-mentioned moving object detection methods for fusing depth information.
The embodiment of the present invention further provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, where the processor implements any one of the above-mentioned moving object detection methods for merging depth information when executing the computer program.
The embodiment of the invention has the following beneficial effects:
the embodiment of the invention provides a moving target detection method, a moving target detection device, a computer readable storage medium and terminal equipment which are integrated with depth information, wherein the method comprises the following steps: acquiring an image to be detected; carrying out depth information extraction processing on an image to be detected to obtain depth information of N pixel points; wherein N > 0; correcting parameters in preset M mixed Gaussian models by using depth information of N pixel points to obtain M moving target detection models; wherein M > 0; performing high contrast retention processing on an image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one; obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models; compared with the prior art that the sensitivity of the Gaussian mixture model is fixed, the method can adaptively adjust the sensitivity of the Gaussian mixture model according to the depth information, improve the detection precision of the moving target area, and accordingly reduce the false detection rate of the moving target.
Drawings
FIG. 1 is a flow chart of a moving object detection method with depth information fused according to a preferred embodiment of the present invention;
FIG. 2 is an image to be detected provided by the present invention;
FIG. 3 is a diagram of an identification image corresponding to a moving target area provided by the present invention;
FIG. 4 is a block diagram of a moving object detecting apparatus with depth information fusion according to a preferred embodiment of the present invention;
fig. 5 is a block diagram of a preferred embodiment of a terminal device provided in the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, 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 any inventive step, are within the scope of the present invention.
An embodiment of the present invention provides a depth information fused moving object detection method, which is a flowchart of a preferred embodiment of a depth information fused moving object detection method provided by the present invention, as shown in fig. 1, and the method includes steps S11 to S15:
step S11, acquiring an image to be detected;
step S12, extracting depth information of the image to be detected to obtain depth information of N pixel points; wherein N > 0;
step S13, correcting parameters in M preset mixed Gaussian models by using depth information of N pixel points to obtain M moving target detection models; wherein M > 0;
s14, performing high contrast reserving processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and step S15, obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models.
Specifically, acquiring and performing depth information extraction processing on an image to be detected to obtain depth information of N pixel points; wherein N is all pixel points in the image to be detected; correcting parameters in preset M mixed Gaussian models by using depth information of N pixel points to correspondingly obtain M moving target detection models; for example, taking the first pixel point as an example, the depth information of the first pixel point is used to correct the parameter in the gaussian mixture model corresponding to the first pixel point, so as to obtain a moving target detection model corresponding to the first pixel point; performing high contrast retention processing on an image to be detected to obtain N high-frequency characteristic values; obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models; compared with the prior art that the sensitivity of the Gaussian mixture model is fixed, the method can adaptively adjust the sensitivity of the Gaussian mixture model according to the depth information, improve the detection precision of the moving target area, and accordingly reduce the false detection rate of the moving target.
In another preferred embodiment, the image sequence to be detected is acquired in real time through electronic equipment with a video recording function; the image sequence to be detected comprises a plurality of images to be detected; the electronic device with the video recording function can be a network camera, a smart phone and a tablet computer, and the specific acquisition mode is not limited in the invention.
In a further preferred embodiment, the depth information comprises N pixel depth values, one pixel depth value corresponding to each pixel point; then, step S12 specifically includes:
and extracting pixel depth values corresponding to N pixel points in the image to be detected through a body sensor, a deep learning algorithm or a binocular vision technology to obtain depth information of the N pixel points.
In another preferred embodiment, step S13 specifically includes:
correcting a variance threshold and a noise variance threshold in a Gaussian mixture model corresponding to each pixel point according to the pixel depth value corresponding to each pixel point to obtain M moving target detection models;
the variance threshold value in the moving target detection model corresponding to the nth pixel point is the variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; the noise variance threshold value in the moving target detection model corresponding to the nth pixel point is the noise variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; n is more than or equal to N and more than 0.
Specifically, each gaussian mixture model includes a variance threshold and a noise variance threshold, taking a first pixel point in the image to be detected as an example, the variance threshold W of the moving object detection model corresponding to the first pixel point is the variance threshold W of the gaussian mixture model corresponding to the first pixel point divided by a pixel depth value d1 corresponding to the first pixel point (which is equivalent to W/d 1); the noise variance threshold P of the moving object detection model corresponding to the first pixel point is obtained by dividing the noise variance threshold P in the gaussian mixture model corresponding to the first pixel point by the pixel depth value d1 corresponding to the first pixel point (which is equivalent to P/d 1). By adjusting the sensitivity of the Gaussian mixture model in the depth information adaptability, a remote moving target can be effectively detected, the detection precision of a moving target area is improved, and therefore the false detection rate and the false alarm rate of the moving target are reduced.
In another preferred embodiment, the step S14 includes the following steps:
filtering the image to be detected by using a preset filter to obtain a blurred image; the method comprises the following steps that pixel points in an image to be detected correspond to pixel points in a blurred image one by one; and obtaining N high-frequency characteristic values according to the image to be detected, the blurred image and a preset high-contrast retention algorithm. The filter may be an average filter, a gaussian filter, or a low-pass filter, but the present invention is not limited thereto.
In another preferred embodiment, the obtaining N high-frequency feature values according to the image to be detected, the blurred image and a preset high contrast retention algorithm specifically includes:
calculating high-frequency characteristic values of N pixel points in the image to be detected according to a calculation formula of a high-contrast retention algorithm to obtain N high-frequency characteristic values; wherein the high contrast preserving algorithm is calculated as follows:
G(n)=T1(n)-T(n)+A;
G(n)is the nth high-frequency characteristic value; t1(n)The pixel value corresponding to the nth pixel point in the blurred image is obtained; t is(n)The pixel value corresponding to the nth pixel point in the image to be detected is obtained; a is a constant greater than 0; among them, A is preferably 128.
In yet another preferred embodiment, after step S15, the method further comprises:
and performing area statistic correction on each pixel point in the moving target area to obtain a corrected moving target area.
In another preferred embodiment, the method performs area statistical correction on any one pixel point in the moving target area by the following steps:
acquiring an m × m neighborhood of the pixel point; wherein, the m-m neighborhood takes the pixel point as a central pixel point; 0< m < w/2; w is the width of the image to be detected;
counting the number num0 of background pixels contained in the m × m neighborhood;
when m x m a is not more than num0, correcting the central pixel point as a background pixel point, otherwise correcting the central pixel point as a moving pixel point; wherein a is more than 0 and less than 1, and a is a preset percentage threshold value.
Specifically, with reference to fig. 2 and 3 and the above embodiments, fig. 2 is an image to be detected, fig. 3 is an identification image corresponding to a moving target region, the number of pixel points in the image to be detected and the identification image corresponding to the moving target region are the same, and the difference is that the pixel points in the identification image corresponding to the moving target region have identification information, and the pixel points are divided into moving pixel points and background pixel points according to the identification information, where the moving pixel points are marked as 1, which is a white region in fig. 3; the background pixel point is marked as 0, namely the black area in fig. 3; taking the first pixel point R in the moving target region as an example, taking a 3 × 3 neighborhood of the first pixel point R, and taking the first pixel point R as a central pixel point of the 3 × 3 neighborhood; counting the number num0 of 0 in the 3 × 3 neighborhood, when 3 × a is not more than num0, correcting the first pixel point R as a background pixel point (namely 0), otherwise, correcting the central pixel point as a moving pixel point (namely 1); the pixel points in the moving target area are corrected, so that the detection precision of the moving target area is further improved, and the false detection rate and the false alarm rate of the moving target are reduced.
The embodiment of the present invention further provides a depth information fused moving object detection apparatus, which can implement all the processes of the depth information fused moving object detection method described in any of the above embodiments, and the functions and implemented technical effects of each module and unit in the apparatus are respectively the same as those of the depth information fused moving object detection method described in the above embodiments and implemented technical correlation, and are not described herein again.
Referring to fig. 4, it is a block diagram of a preferred embodiment of a moving object detection apparatus with depth information fused, provided by the present invention, and the apparatus includes:
the image acquisition module 11 is used for acquiring an image to be detected;
the depth information extraction module 12 is configured to perform depth information extraction processing on the image to be detected to obtain depth information of N pixel points; wherein N > 0;
the model correction module 13 is configured to correct parameters in M preset mixed gaussian models by using depth information of N pixel points, so as to obtain M moving target detection models; wherein M > 0;
the image processing module 14 is configured to perform high contrast preservation processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and the image detection module 15 is configured to obtain a moving target area according to the N high-frequency feature values and the M moving target detection models.
Preferably, the depth information includes N pixel depth values, and each pixel point corresponds to one pixel depth value; then, the depth information extraction module 12 is specifically configured to:
and extracting pixel depth values corresponding to N pixel points in the image to be detected through a body sensor, a deep learning algorithm or a binocular vision technology to obtain depth information of the N pixel points.
Preferably, the model correction module 13 is specifically configured to:
correcting a variance threshold and a noise variance threshold in a Gaussian mixture model corresponding to each pixel point according to the pixel depth value corresponding to each pixel point to obtain M moving target detection models;
the variance threshold value in the moving target detection model corresponding to the nth pixel point is the variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; the noise variance threshold value in the moving target detection model corresponding to the nth pixel point is the noise variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; n is more than or equal to N and more than 0.
Preferably, the image processing module 14 specifically includes:
the first image processing unit is used for carrying out filtering processing on the image to be detected by utilizing a preset filter to obtain a blurred image; the method comprises the following steps that pixel points in an image to be detected correspond to pixel points in a blurred image one by one;
and the second image processing unit is used for obtaining N high-frequency characteristic values according to the image to be detected, the blurred image and a preset high-contrast retention algorithm.
Preferably, the second image processing unit is specifically configured to:
calculating high-frequency characteristic values of N pixel points in the image to be detected according to a calculation formula of a high-contrast retention algorithm to obtain N high-frequency characteristic values; wherein the high contrast preserving algorithm is calculated as follows:
G(n)=T1(n)-T(n)+A;
G(n)is the nth high-frequency characteristic value; t1(n)The pixel value corresponding to the nth pixel point in the blurred image is obtained; t is(n)For the second in the image to be detectedPixel values corresponding to the n pixel points; a is a constant greater than 0.
Preferably, the device further comprises a moving object region correction module; the moving target area correction module is used for performing area statistical correction on each pixel point in the moving target area to obtain a corrected moving target area.
Preferably, the moving object region correction module is specifically configured to perform region statistical correction on any one pixel point in the moving object region through the following steps:
acquiring an m × m neighborhood of the pixel point; wherein, the m-m neighborhood takes the pixel point as a central pixel point; m is an integer greater than 0;
counting the number num0 of background pixels contained in the m × m neighborhood;
when m x m a is not more than num0, correcting the central pixel point as a background pixel point, otherwise correcting the central pixel point as a moving pixel point; wherein a is more than 0 and less than 1.
An embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a stored computer program; when running, the computer program controls a device where the computer-readable storage medium is located to execute the moving object detection method for fusing depth information according to any of the above embodiments.
An embodiment of the present invention further provides a terminal device, as shown in fig. 5, which is a block diagram of a preferred embodiment of the terminal device provided in the present invention, the terminal device includes a processor 10, a memory 20, and a computer program stored in the memory 20 and configured to be executed by the processor 10, and when the computer program is executed, the processor 10 implements the moving object detection method for fusing depth information according to any of the above embodiments.
Preferably, the computer program may be divided into one or more modules/units (e.g., computer program 1, computer program 2, … …) that are stored in the memory 20 and executed by the processor 10 to implement the present invention. The one or more modules/units may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution of the computer program in the air conditioner.
The Processor 10 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, a discrete hardware component, etc., the general purpose Processor may be a microprocessor, or the Processor 10 may be any conventional Processor, the Processor 10 is a control center of the air conditioner, and various interfaces and lines are used to connect various parts of the air conditioner.
The memory 20 mainly includes a program storage area that may store an operating system, an application program required for at least one function, and the like, and a data storage area that may store related data and the like. In addition, the memory 20 may be a high speed random access memory, may also be a non-volatile memory, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card), and the like, or the memory 20 may also be other volatile solid state memory devices.
It should be noted that the terminal device may include, but is not limited to, a processor and a memory, and those skilled in the art will understand that the structural block diagram of fig. 5 is only an example of the terminal device and does not constitute a limitation to the terminal device, and may include more or less components than those shown, or combine some components, or different components.
To sum up, the moving object detection method and apparatus, the computer-readable storage medium, and the terminal device incorporating depth information according to the embodiments of the present invention have the following advantages:
(1) the sensitivity of the Gaussian mixture model is adaptively adjusted according to the depth information, and the detection precision of the moving target area is improved, so that the false detection rate and the false alarm rate of the moving target are reduced.
(2) The detection precision of the moving target area is further improved and the false detection rate and the false alarm rate of the moving target are reduced by correcting the moving target area.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (10)
1. A moving object detection method fusing depth information is characterized by comprising the following steps:
acquiring an image to be detected;
carrying out depth information extraction processing on the image to be detected to obtain depth information of N pixel points; wherein N > 0;
correcting parameters in preset M mixed Gaussian models by using depth information of N pixel points to obtain M moving target detection models; wherein M > 0;
performing high contrast retention processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models.
2. The method as claimed in claim 1, wherein the depth information comprises N pixel depth values, one pixel depth value for each pixel; then, the depth information extraction processing is performed on the image to be detected, so as to obtain the depth information of N pixel points, and the method specifically includes:
and extracting pixel depth values corresponding to N pixel points in the image to be detected through a body sensor, a deep learning algorithm or a binocular vision technology to obtain depth information of the N pixel points.
3. The method for detecting a moving object with depth information fused according to claim 2, wherein the step of correcting parameters in M preset mixed gaussian models by using the depth information of N pixel points to obtain M moving object detection models specifically comprises:
correcting a variance threshold and a noise variance threshold in a Gaussian mixture model corresponding to each pixel point according to the pixel depth value corresponding to each pixel point to obtain M moving target detection models;
the variance threshold value in the moving target detection model corresponding to the nth pixel point is the variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; the noise variance threshold value in the moving target detection model corresponding to the nth pixel point is the noise variance threshold value in the mixed Gaussian model corresponding to the nth pixel point/the pixel depth value corresponding to the nth pixel point; n is more than or equal to N and more than 0.
4. The method for detecting a moving object with fused depth information as claimed in claim 3, wherein said performing high contrast preserving process on the image to be detected to obtain N high frequency feature values specifically comprises:
filtering the image to be detected by using a preset filter to obtain a blurred image; the method comprises the following steps that pixel points in an image to be detected correspond to pixel points in a blurred image one by one;
and obtaining N high-frequency characteristic values according to the image to be detected, the blurred image and a preset high-contrast retention algorithm.
5. The method for detecting a moving object with fused depth information as claimed in claim 4, wherein said obtaining N high frequency feature values according to the image to be detected, the blurred image and a preset high contrast preserving algorithm specifically comprises:
calculating high-frequency characteristic values of N pixel points in the image to be detected according to a calculation formula of a high-contrast retention algorithm to obtain N high-frequency characteristic values; wherein the high contrast preserving algorithm is calculated as follows:
G(n)=T1(n)-T(n)+A;
G(n)is the nth high-frequency characteristic value; t1(n)The pixel value corresponding to the nth pixel point in the blurred image is obtained; t is(n)The pixel value corresponding to the nth pixel point in the image to be detected is obtained; a is a constant greater than 0.
6. The moving object detection method with depth information fused as claimed in claim 1, wherein after said obtaining of moving object regions from N high frequency eigenvalues and M moving object detection models, the method further comprises:
and performing area statistic correction on each pixel point in the moving target area to obtain a corrected moving target area.
7. The method for detecting a moving object fused with depth information according to claim 6, wherein the method performs area statistics correction on any one pixel point in the moving object area by the following steps:
acquiring an m × m neighborhood of the pixel point; wherein, the m-m neighborhood takes the pixel point as a central pixel point; m is an integer greater than 0;
counting the number num0 of background pixels contained in the m × m neighborhood;
when m x m a is not more than num0, correcting the central pixel point as a background pixel point, otherwise correcting the central pixel point as a moving pixel point; wherein a is more than 0 and less than 1.
8. A moving object detecting apparatus that fuses depth information, comprising:
the image acquisition module is used for acquiring an image to be detected;
the depth information extraction module is used for extracting depth information of the image to be detected to obtain depth information of N pixel points; wherein N > 0;
the model correction module is used for correcting parameters in M preset mixed Gaussian models by using the depth information of the N pixel points to obtain M moving target detection models; wherein M > 0;
the image processing module is used for performing high contrast retention processing on the image to be detected to obtain N high-frequency characteristic values; the high-frequency characteristic values correspond to the pixel points one by one;
and the image detection module is used for obtaining a moving target area according to the N high-frequency characteristic values and the M moving target detection models.
9. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises a stored computer program; wherein the computer program controls an apparatus in which the computer-readable storage medium is located to perform the moving object detection method of fusing depth information according to any one of claims 1 to 7 when executed.
10. A terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the moving object detection method fusing depth information according to any one of claims 1 to 7 when executing the computer program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110670125.5A CN113487660A (en) | 2021-06-16 | 2021-06-16 | Depth information fused moving target detection method, device, medium and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110670125.5A CN113487660A (en) | 2021-06-16 | 2021-06-16 | Depth information fused moving target detection method, device, medium and equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113487660A true CN113487660A (en) | 2021-10-08 |
Family
ID=77935515
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110670125.5A Pending CN113487660A (en) | 2021-06-16 | 2021-06-16 | Depth information fused moving target detection method, device, medium and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113487660A (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102316307A (en) * | 2011-08-22 | 2012-01-11 | 安防科技(中国)有限公司 | Road traffic video detection method and apparatus thereof |
CN103700114A (en) * | 2012-09-27 | 2014-04-02 | 中国航天科工集团第二研究院二O七所 | Complex background modeling method based on variable Gaussian mixture number |
CN104952054A (en) * | 2014-03-24 | 2015-09-30 | 联想(北京)有限公司 | Method and device for background modelling |
KR101581586B1 (en) * | 2014-10-13 | 2016-01-11 | 이화여자대학교 산학협력단 | Compensation method for noise of depth image |
US20180322646A1 (en) * | 2016-01-05 | 2018-11-08 | California Institute Of Technology | Gaussian mixture models for temporal depth fusion |
CN110602476A (en) * | 2019-08-08 | 2019-12-20 | 南京航空航天大学 | Hole filling method of Gaussian mixture model based on depth information assistance |
CN110852997A (en) * | 2019-10-24 | 2020-02-28 | 普联技术有限公司 | Dynamic image definition detection method and device, electronic equipment and storage medium |
CN111160340A (en) * | 2019-12-24 | 2020-05-15 | 普联国际有限公司 | Moving target detection method and device, storage medium and terminal equipment |
CN111179302A (en) * | 2019-12-27 | 2020-05-19 | 普联国际有限公司 | Moving target detection method and device, storage medium and terminal equipment |
-
2021
- 2021-06-16 CN CN202110670125.5A patent/CN113487660A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102316307A (en) * | 2011-08-22 | 2012-01-11 | 安防科技(中国)有限公司 | Road traffic video detection method and apparatus thereof |
CN103700114A (en) * | 2012-09-27 | 2014-04-02 | 中国航天科工集团第二研究院二O七所 | Complex background modeling method based on variable Gaussian mixture number |
CN104952054A (en) * | 2014-03-24 | 2015-09-30 | 联想(北京)有限公司 | Method and device for background modelling |
KR101581586B1 (en) * | 2014-10-13 | 2016-01-11 | 이화여자대학교 산학협력단 | Compensation method for noise of depth image |
US20180322646A1 (en) * | 2016-01-05 | 2018-11-08 | California Institute Of Technology | Gaussian mixture models for temporal depth fusion |
CN110602476A (en) * | 2019-08-08 | 2019-12-20 | 南京航空航天大学 | Hole filling method of Gaussian mixture model based on depth information assistance |
CN110852997A (en) * | 2019-10-24 | 2020-02-28 | 普联技术有限公司 | Dynamic image definition detection method and device, electronic equipment and storage medium |
CN111160340A (en) * | 2019-12-24 | 2020-05-15 | 普联国际有限公司 | Moving target detection method and device, storage medium and terminal equipment |
CN111179302A (en) * | 2019-12-27 | 2020-05-19 | 普联国际有限公司 | Moving target detection method and device, storage medium and terminal equipment |
Non-Patent Citations (5)
Title |
---|
CHO, JAECHAN等: "Moving Object Detection Based on Optical Flow Estimation and a Gaussian Mixture Model for Advanced Driver Assistance Systems.Sensors", 《SENSORS》, vol. 19, no. 14, 19 July 2019 (2019-07-19), pages 3217 - 1 * |
R. AZZAM等: "Efficient visual object detection with spatially global Gaussian mixture models and uncertainties", 《JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION》, vol. 36, 30 April 2016 (2016-04-30), pages 90 - 106, XP029440374, DOI: 10.1016/j.jvcir.2015.11.009 * |
周增刚: "基于视频监控的车辆识别技术的研究", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 2013, 15 June 2013 (2013-06-15), pages 138 - 1155 * |
孙挺等: "基于帧间差分和背景差分的运动目标检测算法", 《吉林大学学报(工学版)》, vol. 46, no. 04, 15 July 2016 (2016-07-15), pages 1325 - 1329 * |
郑加敏: "基于自适应混合高斯模型的运动目标检测与跟踪算法研究", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 2018, 15 April 2018 (2018-04-15), pages 138 - 3119 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109299703B (en) | Method and device for carrying out statistics on mouse conditions and image acquisition equipment | |
CN111179302B (en) | Moving target detection method and device, storage medium and terminal equipment | |
CN109064504B (en) | Image processing method, apparatus and computer storage medium | |
CN111210399B (en) | Imaging quality evaluation method, device and equipment | |
CN108280386B (en) | Monitoring scene detection method and device | |
CN110660090B (en) | Subject detection method and apparatus, electronic device, and computer-readable storage medium | |
CN111723644A (en) | Method and system for detecting occlusion of surveillance video | |
CN110796041B (en) | Principal identification method and apparatus, electronic device, and computer-readable storage medium | |
CN111783524A (en) | Scene change detection method and device, storage medium and terminal equipment | |
CN113822223A (en) | Method and device for detecting shielding movement of camera | |
CN111079613A (en) | Gesture recognition method and apparatus, electronic device, and storage medium | |
CN113808135B (en) | Image brightness abnormality detection method, electronic device, and storage medium | |
CN110910429B (en) | Moving target detection method and device, storage medium and terminal equipment | |
CN111160340B (en) | Moving object detection method and device, storage medium and terminal equipment | |
CN111126191B (en) | Iris image acquisition method, iris image acquisition device and storage medium | |
CN110399823B (en) | Subject tracking method and apparatus, electronic device, and computer-readable storage medium | |
CN113052019A (en) | Target tracking method and device, intelligent equipment and computer storage medium | |
CN112581481A (en) | Image processing method and device, electronic equipment and computer readable storage medium | |
CN109598195B (en) | Method and device for processing clear face image based on monitoring video | |
CN108737733B (en) | Information prompting method and device, electronic equipment and computer readable storage medium | |
CN113487660A (en) | Depth information fused moving target detection method, device, medium and equipment | |
CN114037741B (en) | Self-adaptive target detection method and device based on event camera | |
CN111476056A (en) | Target object identification method and device, terminal equipment and computer storage medium | |
CN111027560B (en) | Text detection method and related device | |
CN113822927A (en) | Face detection method, device, medium and equipment suitable for weak-quality images |
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 |