CN103679745A - Moving target detection method and device - Google Patents
Moving target detection method and device Download PDFInfo
- Publication number
- CN103679745A CN103679745A CN201210344459.4A CN201210344459A CN103679745A CN 103679745 A CN103679745 A CN 103679745A CN 201210344459 A CN201210344459 A CN 201210344459A CN 103679745 A CN103679745 A CN 103679745A
- Authority
- CN
- China
- Prior art keywords
- image
- area
- frame image
- current
- current frame
- 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.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a moving target detection method and device for improving moving target detection efficiency and precision. The moving target detection method comprises the following steps: extracting foreground of a current frame image according to the current background; determining whether the pixel number of the foreground is larger than a preset pixel number, and if yes, determining whether the area of a motion region of the current frame image is larger than that of a preset motion region, and if yes, recording the frame number of the image, of which the area of the motion region is larger than that of the preset motion region; determining whether the frame number of the image, of which the area of the motion region is larger than that of the preset motion region in a time period Deltt, is larger than a preset image frame number, and if yes, determining that a moving target exists at present.
Description
Technical field
The present invention relates to technical field of image processing, relate in particular to a kind of moving target detecting method and device.
Background technology
Video monitoring is based on computer video, the video image in monitoring scene to be analyzed, and extracts the image information in monitored scene, forms the warning of moving target, aspect public safety protection and traffic administration, is playing an important role.
Video monitoring system is to use video image acquisition device continuously to take collection video flowing to a certain scene, and the video flowing of collecting is processed, and detects moving target wherein, and moving target is followed the tracks of and subsequent treatment.
In some cases, video monitoring system is calculated without the exact position of the moving target in scene, only need to determine under this scene, whether there is remarkable moving target.
The existing detection method to moving target comprises the following steps:
The first, the current frame video image of obtaining in video flowing is carried out to background modeling, obtain the background image of video image.
The second, utilize the background image obtaining to extract accurately the foreground target in video image, and carry out context update.
Three, the foreground target extracting is followed the tracks of to processing, to judge whether it moves.
A general character of close shot video is exactly that environment is very little, affected by illumination variation very large.The background modeling method of prior art is difficult to adapt to such illumination variation, conventionally can cause modeling failure, foreground extraction mistake.And, the exact position of the foreground target prospect that prior art is extracted, and prospect is followed the tracks of to processing, data processing amount is larger, detection efficiency to moving target is lower, under the larger scene of illumination variation, extract that prospect coordinate position is very difficult accurately, be difficult to obtain accurately moving target or obtain moving target failure.
Prior art, under the larger scene of illumination variation, the detection efficiency of moving target is lower, and is difficult to detect moving target accurately.
Summary of the invention
The embodiment of the present invention provides a kind of moving target detecting method and device, in order to improve the detection efficiency of moving target and accuracy rate.
The moving target detecting method that the embodiment of the present invention provides, comprises the following steps:
A, according to current background, extract the prospect of current frame image;
B, add up the number of pixels of described prospect, judge whether described number of pixels is greater than presetted pixel number, if so, execution step C, otherwise execution step D;
The area of the moving region of C, the relative previous frame image of calculating current frame image, execution step E; Wherein, described previous frame image meets: the number of pixels of prospect is greater than presetted pixel number;
D, current frame image is updated to current background, next frame image is performed step to A as current frame image;
E, judge whether the area of the moving region of current frame image is greater than the area in predetermined movement region, if so, execution step F, otherwise execution step D;
F, judgement add up first number of image frames and start the time t that timing is experienced from zero clearing each time, whether be greater than Preset Time interval of delta t, if not, execution step G, if, the area that meets moving region of accumulative total is greater than to the number of image frames zero clearing of the area in predetermined movement region, execution step D; Wherein, the area that the described image adding up in number of image frames meets moving region is greater than the area in predetermined movement region;
The area that G, accumulative total meet moving region is greater than the number of image frames of the area in predetermined movement region;
Whether the number of image frames of H, judgement accumulative total is greater than default number of image frames, if so, and execution step I, otherwise execution step D.
I, determine the current moving target that exists.
The moving object detection device that the embodiment of the present invention provides, comprising:
Foreground extraction unit, for extracting the prospect of current frame image according to current background;
Pixels statistics unit, for adding up the number of pixels of described prospect;
The first judging unit, for judging whether the number of pixels of the prospect of described pixels statistics unit statistics is greater than presetted pixel number;
Moving region computing unit, for calculating the area of the moving region of the relative former frame image of current frame image, wherein, described former frame image meets: the number of pixels of the prospect of former frame image is greater than described presetted pixel number;
The second judging unit, for judging whether the moving region area of the present frame that moving region computing unit calculates is greater than predetermined movement region area;
The 3rd judging unit, for judging whether adding up first number of image frames from zero clearing each time starts the time t that timing is experienced, be greater than Preset Time interval of delta t; Wherein, the area that the described image adding up in number of image frames meets moving region is greater than the area in predetermined movement region;
Number of image frames zero clearing unit, for when the 3rd judging unit judgement t is greater than Δ t, is greater than the area that meets moving region of accumulative total the number of image frames zero clearing of the area in predetermined movement region;
Record cell, for when the 3rd judging unit judges that described t is less than Δ t, the area that totally meets moving region is greater than the number of image frames of the area in predetermined movement region;
The 4th judging unit, for judging whether the number of image frames of recording unit records is greater than default number of image frames;
Moving target determining unit, while being greater than default number of image frames for the number of image frames when the 4th judging unit judgement record, determines the current moving target that exists;
Context update unit, for being updated to current background by current frame image.
The embodiment of the present invention, according to the prospect of current background extraction current frame image; Whether the number of pixels that judges prospect is greater than presetted pixel number, if, judge whether the area of the moving region of current frame image is greater than the area in predetermined movement region, if so, the area that record meets moving region is greater than the number of image frames of the area in predetermined movement region; Whether the number of image frames that the area that judgement meets moving region in Δ t is greater than the area in predetermined movement region is greater than default number of image frames, if, determine the current moving target that exists, above-mentioned in all backgrounds of upgrading the determination result is NO in the situation that, next frame image is continued to above-mentioned steps as current frame image, wherein, described Δ t is for starting from adding up first image the time that timing is experienced.The present invention is without each two field picture is carried out to foreground extraction accurately, and saved without the area of determining the moving region of each two field picture the expense of calculating, the detection efficiency of raising to moving target, because the present invention does not need to determine prospect coordinate accurately, when light changes greatly, still can detect target image accurately, improve and detect target image and accuracy rate.
Accompanying drawing explanation
The moving target detecting method overall procedure schematic diagram that Fig. 1 provides for the embodiment of the present invention;
The background modeling method schematic flow sheet that Fig. 2 provides for the embodiment of the present invention;
The foreground image acquisition methods schematic flow sheet that Fig. 3 provides for the embodiment of the present invention;
The moving target detecting method idiographic flow schematic diagram that Fig. 4 provides for the embodiment of the present invention;
The moving object detection apparatus structure schematic diagram that Fig. 5 provides for the embodiment of the present invention;
The moving object detection apparatus structure schematic diagram that Fig. 6 provides for the embodiment of the present invention.
Embodiment
The embodiment of the present invention provides a kind of moving target detecting method and device, in order to improve detection efficiency and the accuracy rate of moving target.
Because foreground target is generally all more greatly or more significant target, the moving target detecting method that the embodiment of the present invention provides, each frame video image is not extracted to prospect accurately, only need to judge whether it has the prospect in larger region, if there is larger foreground area, judged whether again significant moving target, otherwise upgrade background, continued to judge whether next frame video image has the prospect in larger region.If there is significant moving target, Output rusults, otherwise upgrade background, continue to judge whether next frame video image has the prospect in larger region.
The present invention only judges according to the size of foreground area, each two field picture is not followed the tracks of, improved the detection efficiency of moving target, and the present invention does not need to extract prospect accurately, when under the larger scene of illumination variation, moving target can be detected, improve the accuracy rate to moving object detection.
Below by accompanying drawing, illustrate the technical scheme that the embodiment of the present invention provides.
Referring to Fig. 1, the moving target detecting method that the embodiment of the present invention provides, comprises the following steps:
S10, according to current background, extract the prospect of current frame image, execution step S11.
S11, add up the number of pixels of described prospect.
S12, judge whether the number of pixels of described prospect is greater than presetted pixel number, if so, execution step S13, otherwise execution step S14;
Presetted pixel number can be determined according to image size, the 20% corresponding number of pixels of generally getting image size (i.e. figure image width * figure image height).
The area of the moving region of S13, the relative former frame image of calculating current frame image, execution step 15; Wherein, described former frame image meets: the number of pixels of prospect is greater than presetted pixel number;
The calculating of the area of the moving region of the relative former frame image of described current frame image, is specially:
Calculate the optical flow field of current frame image and former frame image, obtain the area of the moving region of current frame image, execution step S15; Wherein, described former frame image meets: the number of pixels of prospect is greater than described presetted pixel number.
S14, current frame image is updated to current background, using next frame image as present frame, execution step S10.
S15, judge whether the area of the moving region of current frame image is greater than the area in predetermined movement region, if so, execution step S16, otherwise execution step S14.
S16, judgement add up first number of image frames and start the time t that timing is experienced from zero clearing each time, whether be greater than Preset Time interval of delta t, if not, execution step S17, if, execution step S20, wherein, the area that the described image adding up in number of image frames meets moving region is greater than the area in predetermined movement region.
Described predetermined movement region is determined according to image size, is generally got 10% of image size (i.e. figure image width * figure image height).
The area that S17, accumulative total meet moving region is greater than the number of image frames of the area in predetermined movement region.
Whether the number of image frames of S18, judgement accumulative total is greater than default number of image frames, if so, and execution step S20, otherwise execution step S14.
S19, determine the current moving target that exists.
After execution step S19, the method also comprises: determine that described moving target leaves.
Particularly, perform step successively A and B, and meet after number of pixels is less than presetted pixel number in step B judged result, also comprise: judgement determine current there is moving target after, whether the judged result that the number of pixels that meets current frame image is less than presetted pixel number occurs first, if, determine that described moving target leaves, the area that meets moving region of accumulative total is greater than to the frame number zero clearing of picture frame of the area in predetermined movement region, continue execution step D, start to detect next moving target; Otherwise continue to perform step successively A and B.
It should be noted that, before definite described moving target leaves, without upgrading background, only need to upgrade prospect, according to the number of pixels of prospect, judge whether moving target leaves.
Preferably, when determine current exist moving target or definite moving target to leave after, carry out and report to the police, be about to moving target output.
Be not more than to the update cycle of current frame image the processing of each two field picture time used, can guarantee that to the processing of each two field picture be all the processing to current frame image, when next two field picture arrives, former frame image is disposed.Therefore, without next frame picture delay is processed.
Described default image number can be set according to actual conditions, chooses calculated amount less and detect moving target empirical value more accurately.When default image number is larger, calculatively meet moving region to be greater than the image of predeterminable area more, the False Rate of the moving target obtaining is lower, but calculated amount is larger, therefore selects a suitable empirical value comparatively reasonable.
Because moving target is a continuous motion process, when there is no moving target in Preset Time Δ t, also there are some erroneous judgements, when meeting moving region area as existed and being greater than the image of area in predetermined movement region, need to be the number of image frames zero clearing of record, not affect the accurate judgement to subsequent motion target detection.
Preferably, described image is video image, and described video image can be the real time data gathering from video capture device that moving object detection device obtains, and can be also the video flowing that moving object detection device is preserved in advance.
Before execution step S10, the method also comprises: according to the image of default frame number, carry out background modeling, the background obtaining is described current background.
Referring to Fig. 2, according to the image of default frame number, carry out background modeling, the detailed process that the background obtaining is described current background comprises the following steps:
S21, the gray scale summation of continuous some two field pictures is averaged, obtain cumulative mean frame.
Preferably, the span of described continuous some two field pictures can be 5 ~ 50 frames.
For example, preferably, can the summation of the Y component under YUV color space average to front 25 frames in video image, obtain cumulative mean frame (or also claiming average frame).
S22, resulting cumulative mean frame is carried out to histogram specification process and to obtain background image.
Described histogram specification is to remove the pending image of regulationization with a default histogram, makes the histogram distribution of image approach default histogram.
Referring to Fig. 3, in step S10, the extracting method of the prospect of current frame image comprises the steps:
S31, current frame image is carried out to histogram specification processing.
S32, the background image obtaining in Fig. 2 and the current frame image after histogram specification is processed are carried out to absolute Difference Calculation, obtain foreground image.
Described absolute difference, carries out subtraction to the gray scale of pixel corresponding to two width images and then takes absolute value.The present invention carries out subtraction by the gray scale of background image and pixel corresponding to the current frame image after histogram specification is processed then to take absolute value.
The present invention is to process by the mode of histogram specification to the modeling process of background, algorithm operation quantity is little, when the moving region of judgement current frame image is less than predetermined movement region, do not need current frame image to carry out optical flow computation relatively consuming time, method is easy to meet the requirement of real-time.
Above-mentioned is only the size in the region of foreground image to the obtain manner of foreground image, does not obtain the concrete coordinate figure of foreground image.Save a large amount of calculated amount, improved the efficiency that detects moving target.
Introduce the process of the area of the moving region of the relative former frame image of calculating current frame image described in step S13 in Fig. 1 below.
Particularly, by calculating current frame image and the optical flow field of former frame image, determine the area of the moving region of the relative former frame image of prior image frame.
First in order to reduce the impact of noise of video image on subsequent optical Flow Field Calculation, the image of present frame and former frame optical flow field to be calculated is carried out to Gaussian smoothing.
Light stream has reflected the image change causing due to motion in a certain time interval.Because the viewed brightness of any object point is at short notice invariable, the formula of image overall light stream, as (1) formula, represents that the pixel of (x, y) position moves to (x+dx, y+dy) at dt in the time:
f(x+dx,y+dy,t+dt)=f(x,y,t) (1)
If dynamic image is expressed as to the function of room and time, and is carried out Taylor expansion and can be obtained, wherein
represent higher order term:
Dx in most situation, dy, dt is very little, and higher order term is negligible, and comprehensive (1) and (2) two formulas can obtain (3) formula, u wherein, v is image optical flow field to be asked, and has represented the movement velocity of object:
Method (the Two-Frame Motion Estimation Based on Polynomial Expansion based on polygon expansion that the present invention adopts Gunnar Farneback to propose, in Proceedings of the 13th Scandinavian Conference on Image Analysis, LNCS2749, (Gothenburg, Sweden), pp.363 – 370, June-July 2003), and its realization has been carried out to the optimal design of floating number fixed point, make its computing not relate to floating number, more easily be transplanted on embedded device, travelling speed is also faster.
The larger prospect flase drop that the one large effect of calculating optical flow field can suddenly change light exactly, scene changes etc. causes is got rid of, the practicality of increase method.What use that the method for optical flow field pays close attention to due to the present invention is remarkable moving target, some frames of being separated by between two two field pictures of obtained calculating optical flow field, so more easily highlight the motion amplitude of moving target, moving target more easily detected, and the calculated amount of method is declined, more easily meet the requirement of real-time.
Substantially be all directly proportional to the size of processed image the operation time of various Processing Algorithm in addition, in order to reduce that the operand of image is improved to the detection efficiency to moving target, preferably, before present image is carried out to background modeling, also comprise: current frame image is carried out to compression of images or intercepting makes it image size in image magnitude range given in advance.
Described image magnitude range can adopt QCIF(Quarter Common Intermediate Format) form.So-called QCIF form is conventional standardized images form.In protocol family H.323, stipulated the standard acquisition resolution of video capture device, QCIF=176 * 144 pixel.
The technical scheme below the detailed description embodiment of the present invention being provided.
Referring to Fig. 4, the moving target detecting method that the embodiment of the present invention provides, specifically comprises the following steps:
S101, obtain some two field pictures of default frame number, as front 25 frames.
S102, according to default image size, respectively described front 25 two field pictures are compressed or intercepted.
S103, the Y component (gray scale) under the YUV color space of described front 25 two field pictures is averaged, obtain cumulative mean frame, resulting cumulative mean frame is carried out to histogram specification and process and obtain background image, described cumulative mean frame is also average frame.
When the 26th two field picture arrives, the 26th two field picture is current frame image, is the processing procedure to the 26th two field picture below.
S104, according to default image size, current frame image is compressed or intercept.
The prospect of S105, the current frame image according to described background image extraction compression or after intercepting.
S106, add up the number of pixels of the prospect of described current frame image.
S107, judge whether the number of pixels of described prospect is greater than presetted pixel number, if so, execution step S108, otherwise, execution step S109.
The optical flow field of S108, calculating former frame image (the 25th two field picture) and current frame image (the 26th two field picture), obtains current frame image with respect to the moving region of former frame image.
S109, judgement determine current there is moving target after, whether the judged result that the number of pixels that meets current frame image is less than presetted pixel number occur first, if so, execution step S115, otherwise execution step S104.
S110, judge whether the area of the moving region of current frame image is greater than predetermined movement region area, if execution step S112, otherwise execution step S116.
S111, judgement add up first number of image frames and start the time t that timing is experienced from zero clearing each time, whether are greater than Preset Time interval of delta t, and if NO, execution step S112, performs step S116 if YES.
The area that S112, accumulative total meet moving region is greater than the number of image frames of the area in predetermined movement region.
Whether the number of image frames of S113, judgement accumulative total is greater than default number of image frames, if so, and execution step S114, otherwise execution step S116.
S114, determine the current moving target that exists, execution step S104.
S115, determine that the moving target occurring last time leaves, execution step S117.
S116, current frame image is updated to current background, next frame image is performed step to S104 as current frame image.
S117, accumulative total met to the picture frame zero clearing that moving region area is greater than predetermined movement region area, execution step S116.
Introduce the moving object detection device that the embodiment of the present invention provides below, referring to Fig. 5, comprising:
Foreground extraction unit 11, for extracting the prospect of current frame image according to current background;
The first judging unit 13, for judging whether the number of pixels of the prospect of pixels statistics unit 12 statistics is greater than presetted pixel number;
Moving region computing unit 14, for calculating the area of the moving region of the relative former frame image of current frame image, wherein, described former frame image meets: the number of pixels of the prospect of former frame image is greater than described presetted pixel number;
The second judging unit 15, for judging whether the area of the moving region of the present frame that moving region computing unit calculates is greater than the area in predetermined movement region;
The 3rd judging unit 16, the number of image frames that is greater than the area in predetermined movement region for judging from totally meeting first the area of moving region starts the time t that timing is experienced, and whether is greater than Preset Time interval of delta t;
Number of image frames zero clearing unit 17, for when the 3rd judging unit judgement t is greater than Δ t, is greater than the area that meets moving region of accumulative total the number of image frames zero clearing of the area in predetermined movement region;
The 4th judging unit 19, for judging whether the number of image frames of recording unit records is greater than default number of image frames;
Moving target determining unit 20, while being greater than default number of image frames for the number of image frames when the 4th judging unit judgement record, determines the current moving target that exists;
Preferably, referring to Fig. 6, described device also comprises:
The 5th judging unit 22, for judgement determine current there is moving target after, whether the judged result that the number of pixels that meets current frame image is less than presetted pixel number occurs first;
Moving target leaves determining unit 23, when the judged result that is less than presetted pixel number for the number of pixels at the 5th judging unit 22 judgement current frame images occurs first, determines that described moving target leaves;
Number of image frames zero clearing unit 17 also for, after definite moving target leaves, the area that meets moving region of accumulative total is greater than to the frame number zero clearing of picture frame of the area in predetermined movement region.
Preferably, foreground extraction unit 11 specifically for: current frame image is carried out to histogram specification processing; Current frame image and described current background after histogram specification is processed are carried out to absolute Difference Calculation, obtain the prospect of current frame image.
Referring to Fig. 6, this device also comprises:
Referring to Fig. 6, this device also comprises:
In sum, the embodiment of the present invention is extracted the prospect of current frame image according to current background; Whether the number of pixels that judges prospect is greater than presetted pixel number, if, judge whether the area of the moving region of current frame image is greater than the area in predetermined movement region, if so, the area that record meets moving region is greater than the number of image frames of the area in predetermined movement region; Whether the number of image frames that the area that judgement meets moving region in Δ t is greater than the area in predetermined movement region is greater than default number of image frames, if, determine the current moving target that exists, above-mentioned in all backgrounds of upgrading the determination result is NO in the situation that, next frame image is continued to above-mentioned steps as current frame image, wherein, described Δ t is for to start from first record image the time that timing is experienced.The present invention is without each two field picture is carried out to foreground extraction accurately, and saved without the area of determining the moving region of each two field picture the expense of calculating, the detection efficiency of raising to moving target, because the present invention does not need to determine prospect coordinate accurately, when light changes greatly, still can detect target image accurately, improve and detect target image and accuracy rate.
Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt complete hardware implementation example, implement software example or in conjunction with the form of the embodiment of software and hardware aspect completely.And the present invention can adopt the form that wherein includes the upper computer program of implementing of computer-usable storage medium (including but not limited to magnetic disk memory and optical memory etc.) of computer usable program code at one or more.
The present invention is with reference to describing according to process flow diagram and/or the block scheme of the method for the embodiment of the present invention, equipment (system) and computer program.Should understand can be in computer program instructions realization flow figure and/or block scheme each flow process and/or the flow process in square frame and process flow diagram and/or block scheme and/or the combination of square frame.Can provide these computer program instructions to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, the instruction of carrying out by the processor of computing machine or other programmable data processing device is produced for realizing the device in the function of flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame appointments.
These computer program instructions also can be stored in energy vectoring computer or the computer-readable memory of other programmable data processing device with ad hoc fashion work, the instruction that makes to be stored in this computer-readable memory produces the manufacture that comprises command device, and this command device is realized the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame.
These computer program instructions also can be loaded in computing machine or other programmable data processing device, make to carry out sequence of operations step to produce computer implemented processing on computing machine or other programmable devices, thereby the instruction of carrying out is provided for realizing the step of the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame on computing machine or other programmable devices.
Obviously, those skilled in the art can carry out various changes and modification and not depart from the spirit and scope of the present invention the present invention.Like this, if within of the present invention these are revised and modification belongs to the scope of the claims in the present invention and equivalent technologies thereof, the present invention is also intended to comprise these changes and modification interior.
Claims (10)
1. a moving target detecting method, is characterized in that, the method comprises the following steps:
A, according to current background, extract the prospect of current frame image;
B, add up the number of pixels of described prospect, judge whether described number of pixels is greater than presetted pixel number, if so, execution step C, otherwise execution step D;
The area of the moving region of C, the relative previous frame image of calculating current frame image, execution step E; Wherein, described previous frame image meets: the number of pixels of prospect is greater than presetted pixel number;
D, current frame image is updated to current background, next frame image is performed step to A as current frame image;
E, judge whether the area of the moving region of current frame image is greater than the area in predetermined movement region, if so, execution step F, otherwise execution step D;
F, judgement add up first number of image frames and start the time t that timing is experienced from number of image frames zero clearing each time, whether be greater than Preset Time interval of delta t, if not, execution step G, if, the area that meets moving region of accumulative total is greater than to the number of image frames zero clearing of the area in predetermined movement region, execution step D; Wherein, the area that the described image adding up in number of image frames meets moving region is greater than the area in predetermined movement region;
The area that G, accumulative total meet moving region is greater than the number of image frames of the area in predetermined movement region;
Whether the number of image frames of H, judgement accumulative total is greater than default number of image frames, if so, and execution step I, otherwise execution step D;
I, determine the current moving target that exists.
2. method according to claim 1, is characterized in that, after execution step I, the method also comprises:
Perform step successively A and B, and meet after number of pixels is less than presetted pixel number in step B judged result, also comprise: judgement determine current there is moving target after, whether the judged result that the number of pixels that meets current frame image is less than presetted pixel number occurs first, if, determine that described moving target leaves, the area that meets moving region of accumulative total is greater than to the frame number zero clearing of picture frame of the area in predetermined movement region, continue execution step D; Otherwise continue to perform step successively A and B.
3. method according to claim 1, is characterized in that, steps A is extracted the prospect of current frame image according to current background, be specially:
Current frame image is carried out to histogram specification processing;
Current frame image and described current background after histogram specification is processed are carried out to absolute Difference Calculation, obtain the prospect of current frame image.
4. method according to claim 1, is characterized in that, before execution step A, the method also comprises: according to the image of default frame number, carry out background modeling, the background obtaining is described current background; Be specially: the gray scale summation to front some two field pictures is averaged, and obtains average frame, and described average frame is carried out to histogram specification processing, obtains current background.
5. method according to claim 1, is characterized in that, in steps A, before extracting the prospect of current frame image, also comprises: the size according to default image is carried out compression of images or intercepting to current frame image.
6. a moving object detection device, is characterized in that, this device comprises:
Foreground extraction unit, for extracting the prospect of current frame image according to current background;
Pixels statistics unit, for adding up the number of pixels of described prospect;
The first judging unit, for judging whether the number of pixels of the prospect of described pixels statistics unit statistics is greater than presetted pixel number;
Moving region computing unit, for calculating the area of the moving region of the relative former frame image of current frame image, wherein, described former frame image meets: the number of pixels of the prospect of former frame image is greater than described presetted pixel number;
The second judging unit, for judging whether the moving region area of the present frame that moving region computing unit calculates is greater than predetermined movement region area;
The 3rd judging unit, the number of image frames that is greater than the area in predetermined movement region for judging from totally meeting first the area of moving region starts the time t that timing is experienced, and whether is greater than Preset Time interval of delta t;
Number of image frames zero clearing unit, for when the 3rd judging unit judgement t is greater than Δ t, is greater than the area that meets moving region of accumulative total the number of image frames zero clearing of the area in predetermined movement region;
Record cell, for when the 3rd judging unit judges that described t is less than Δ t, the area that totally meets moving region is greater than the number of image frames of the area in predetermined movement region;
The 4th judging unit, for judging whether the number of image frames of recording unit records is greater than default number of image frames;
Moving target determining unit, while being greater than default number of image frames for the number of image frames when the 4th judging unit judgement record, determines the current moving target that exists;
Context update unit, for being updated to current background by current frame image.
7. device according to claim 6, is characterized in that, also comprises:
The 5th judging unit, for judgement determine current there is moving target after, whether the judged result that the number of pixels that meets current frame image is less than presetted pixel number occurs first;
Moving target leaves determining unit, when the judged result that is less than presetted pixel number for the number of pixels at the 5th judging unit judgement current frame image occurs first, determines that described moving target leaves;
Described number of image frames zero clearing unit, also for after leaving at definite described moving target, is greater than the area that meets moving region of accumulative total the frame number zero clearing of picture frame of the area in predetermined movement region.
8. device according to claim 6, is characterized in that, described foreground extraction unit specifically for: current frame image is carried out to histogram specification processing; Current frame image and described current background after histogram specification is processed are carried out to absolute Difference Calculation, obtain the prospect of current frame image.
9. device according to claim 6, is characterized in that, this device also comprises:
Background modeling unit, for carrying out background modeling according to the image of default frame number, the background obtaining is described current background; Specifically for, the gray scale summation of continuously front some two field pictures is averaged, obtain average frame, described average frame is carried out to histogram specification processing, obtain current background.
10. device according to claim 6, is characterized in that, this device also comprises:
Picture editting unit, for carrying out compression of images or intercepting according to the size of default image to current frame image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210344459.4A CN103679745B (en) | 2012-09-17 | 2012-09-17 | A kind of moving target detecting method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210344459.4A CN103679745B (en) | 2012-09-17 | 2012-09-17 | A kind of moving target detecting method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103679745A true CN103679745A (en) | 2014-03-26 |
CN103679745B CN103679745B (en) | 2016-08-17 |
Family
ID=50317195
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210344459.4A Active CN103679745B (en) | 2012-09-17 | 2012-09-17 | A kind of moving target detecting method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103679745B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106710348A (en) * | 2016-12-20 | 2017-05-24 | 江苏前景信息科技有限公司 | Civil air defense interactive experience method and system |
CN107566723A (en) * | 2017-09-13 | 2018-01-09 | 维沃移动通信有限公司 | A kind of image pickup method, mobile terminal and computer-readable recording medium |
CN109255360A (en) * | 2017-07-12 | 2019-01-22 | 杭州海康威视数字技术股份有限公司 | A kind of objective classification method, apparatus and system |
CN109379594A (en) * | 2018-10-31 | 2019-02-22 | 北京佳讯飞鸿电气股份有限公司 | Video coding compression method, device, equipment and medium |
CN111191556A (en) * | 2019-12-25 | 2020-05-22 | 杭州宇泛智能科技有限公司 | Face recognition method and device and electronic equipment |
CN111310733A (en) * | 2020-03-19 | 2020-06-19 | 成都云盯科技有限公司 | Method, device and equipment for detecting personnel entering and exiting based on monitoring video |
CN111308969A (en) * | 2020-01-16 | 2020-06-19 | 浙江大学 | Carrier motion mode discrimination method based on time domain differential characteristics |
CN111832357A (en) * | 2019-04-19 | 2020-10-27 | 苏州涟漪信息科技有限公司 | Mobile event detection method and device |
CN112037266A (en) * | 2020-11-05 | 2020-12-04 | 北京软通智慧城市科技有限公司 | Falling object identification method and device, terminal equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1449186A (en) * | 2003-04-03 | 2003-10-15 | 上海交通大学 | Abnormal object automatic finding and tracking video camera system |
EP1667080A1 (en) * | 2001-10-17 | 2006-06-07 | Biodentity Systems Corporation | Face imaging system for recordal and automated identity confirmation |
CN101261681A (en) * | 2008-03-31 | 2008-09-10 | 北京中星微电子有限公司 | Road image extraction method and device in intelligent video monitoring |
CN102222349A (en) * | 2011-07-04 | 2011-10-19 | 江苏大学 | Prospect frame detecting method based on edge model |
WO2012019417A1 (en) * | 2010-08-10 | 2012-02-16 | 中国科学院自动化研究所 | Device, system and method for online video condensation |
-
2012
- 2012-09-17 CN CN201210344459.4A patent/CN103679745B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1667080A1 (en) * | 2001-10-17 | 2006-06-07 | Biodentity Systems Corporation | Face imaging system for recordal and automated identity confirmation |
CN1449186A (en) * | 2003-04-03 | 2003-10-15 | 上海交通大学 | Abnormal object automatic finding and tracking video camera system |
CN101261681A (en) * | 2008-03-31 | 2008-09-10 | 北京中星微电子有限公司 | Road image extraction method and device in intelligent video monitoring |
WO2012019417A1 (en) * | 2010-08-10 | 2012-02-16 | 中国科学院自动化研究所 | Device, system and method for online video condensation |
CN102222349A (en) * | 2011-07-04 | 2011-10-19 | 江苏大学 | Prospect frame detecting method based on edge model |
Non-Patent Citations (4)
Title |
---|
JOST KOLLER,MARTIN ULMKE: "Data fusion for ground moving target tracking", 《AEROSPACE SCIENCE AND TECHNOLOGY》 * |
张宇 等: "一种应用于嵌入式图像监控系统的运动检测方法", 《2009年中国高校通信类院系学术研讨会论文集》 * |
潘石柱 等: "基于自适应背景的实时运动物体检测", 《计算机应用》 * |
王笑雨: "运动目标检测与跟踪系统设计", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106710348A (en) * | 2016-12-20 | 2017-05-24 | 江苏前景信息科技有限公司 | Civil air defense interactive experience method and system |
CN109255360B (en) * | 2017-07-12 | 2021-04-02 | 杭州海康威视数字技术股份有限公司 | Target classification method, device and system |
CN109255360A (en) * | 2017-07-12 | 2019-01-22 | 杭州海康威视数字技术股份有限公司 | A kind of objective classification method, apparatus and system |
CN107566723A (en) * | 2017-09-13 | 2018-01-09 | 维沃移动通信有限公司 | A kind of image pickup method, mobile terminal and computer-readable recording medium |
CN107566723B (en) * | 2017-09-13 | 2019-11-19 | 维沃移动通信有限公司 | A kind of image pickup method, mobile terminal and computer readable storage medium |
CN109379594A (en) * | 2018-10-31 | 2019-02-22 | 北京佳讯飞鸿电气股份有限公司 | Video coding compression method, device, equipment and medium |
CN109379594B (en) * | 2018-10-31 | 2022-07-19 | 北京佳讯飞鸿电气股份有限公司 | Video coding compression method, device, equipment and medium |
CN111832357A (en) * | 2019-04-19 | 2020-10-27 | 苏州涟漪信息科技有限公司 | Mobile event detection method and device |
CN111191556A (en) * | 2019-12-25 | 2020-05-22 | 杭州宇泛智能科技有限公司 | Face recognition method and device and electronic equipment |
CN111308969A (en) * | 2020-01-16 | 2020-06-19 | 浙江大学 | Carrier motion mode discrimination method based on time domain differential characteristics |
CN111310733A (en) * | 2020-03-19 | 2020-06-19 | 成都云盯科技有限公司 | Method, device and equipment for detecting personnel entering and exiting based on monitoring video |
CN111310733B (en) * | 2020-03-19 | 2023-08-22 | 成都云盯科技有限公司 | Personnel access detection method, device and equipment based on monitoring video |
CN112037266A (en) * | 2020-11-05 | 2020-12-04 | 北京软通智慧城市科技有限公司 | Falling object identification method and device, terminal equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN103679745B (en) | 2016-08-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103679745A (en) | Moving target detection method and device | |
EP2858008B1 (en) | Target detecting method and system | |
CN102982537B (en) | A kind of method and system detecting scene change | |
CN103678299A (en) | Method and device for monitoring video abstract | |
EP2927871A1 (en) | Method and device for calculating number of pedestrians and crowd movement directions | |
US20200027240A1 (en) | Pedestrian Tracking Method and Electronic Device | |
CN101022505A (en) | Mobile target in complex background automatic testing method and device | |
CN109711256B (en) | Low-altitude complex background unmanned aerial vehicle target detection method | |
CN102915545A (en) | OpenCV(open source computer vision library)-based video target tracking algorithm | |
CN103810717A (en) | Human behavior detection method and device | |
KR101348680B1 (en) | Target acquisition method for video tracker, and target acquisition apparatus using the same | |
CN105574894A (en) | Method and system for screening moving object feature point tracking results | |
CN111353496B (en) | Real-time detection method for infrared dim targets | |
Vlaminck et al. | Obstacle detection for pedestrians with a visual impairment based on 3D imaging | |
CN111160203A (en) | Loitering and lingering behavior analysis method based on head and shoulder model and IOU tracking | |
CN105469054B (en) | The model building method of normal behaviour and the detection method of abnormal behaviour | |
CN115527158A (en) | Method and device for detecting abnormal behaviors of personnel based on video monitoring | |
CN104683765A (en) | Video concentration method based on mobile object detection | |
CN102222349A (en) | Prospect frame detecting method based on edge model | |
CN108830204A (en) | The method for detecting abnormality in the monitor video of target | |
CN101916380B (en) | Video-based device and method for detecting smog | |
CN113256683B (en) | Target tracking method and related equipment | |
CN101877135B (en) | Moving target detecting method based on background reconstruction | |
CN103905826A (en) | Self-adaptation global motion estimation method | |
CN107067411B (en) | Mean-shift tracking method combined with dense features |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |