CN104021576A - Method and system for tracking moving objects in scene - Google Patents
Method and system for tracking moving objects in scene Download PDFInfo
- Publication number
- CN104021576A CN104021576A CN201410274348.XA CN201410274348A CN104021576A CN 104021576 A CN104021576 A CN 104021576A CN 201410274348 A CN201410274348 A CN 201410274348A CN 104021576 A CN104021576 A CN 104021576A
- Authority
- CN
- China
- Prior art keywords
- mobile object
- video signal
- digital video
- scene
- 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.)
- Pending
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method and system for tracking moving objects in a scene. The method includes the following steps that a digitalized video signal corresponding to the scene is acquired; at least one moving object is recognized from the digital video signal of the scene; after at least one moving object is recognized, the moving object is found in the digital video signal; in a period of preset time after the moving object is recognized, the digital video signal continues be positioned; a score is allocated to each tracked moving object according to the time that the tracked moving object can still be located in the digital video signal after the preset time; the overall score of the scene is acquired according to the score of each tracked moving object so that whether the overall score of a user meets a preset standard or not can be displayed. According to the method, the moving characteristic of a video or an image sequence is analyzed according to whether the absolute value of the brightness difference of two frames of images is larger than a threshold value or not, and whether a moving object exists in the images or not can be determined; most interference can be eliminated.
Description
Technical field
The present invention relates to digital image processing techniques, relate in particular to a kind of method and system of following the trail of mobile object in scene.
Background technology
Video monitoring system has been installed in power plant, transformer station by some Utilities Electric Co. at present, can realize supervision field apparatus, controls the functions such as remote camera action.But these video monitoring systems only have video monitoring function, there is no video image identification function, in order fully to send out the function of video monitoring system, the reason of the on-the-spot alarm that has an accident of judgement more accurately, should adopt remote digital video monitoring and digital image acquisition system, to realize the image recognition of equipment alarm, for detecting, accident provides new means, for crash analysis provides reliable foundation.
In scene, following the trail of mobile object is a problem that appears at digital image processing field.In some applications, need the object of tracking relatively very fast by scene.And in other application, need to follow the trail of the object moving relative to slowly, they can stay the longer time in scene.
First, some little foreign matter distant surveillances are more difficult, and can only take several pixel numeral video frame capture.Secondly in power system transmission line, conventionally have the vegetation of rustling or the noise of of short duration flicker sunlight.Certainly, also having some rudimentary intrinsic noises for any digital imaging apparatus of capturing scenes.Finally, scene also comprises the fastish scene of passing through of our uninterested mobile objects (as bird or aircraft) sometimes.
Summary of the invention
For the weak point existing in above-mentioned technology, the invention provides one in the situation that multiple objects to be detected are deposited simultaneously, can effectively reduce the method and system of following the trail of mobile object in scene of detection frequency operand and same moving target.
For achieving the above object, the invention provides a kind of method of following the trail of mobile object in scene, comprise following step:
Step 1, obtains the digital video signal corresponding with this scene;
Step 2 is identified at least one mobile object in the digital video signal of scene;
Step 3 after at least one mobile object of identification, is found out this mobile object in digital video signal;
Step 4, one period of schedule time after mobile object is identified, continues to be positioned digital video signal;
Step 5, can continue to be positioned at time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object, and the time, longer mark was higher;
Step 6, obtains the overall score of scene in order to show whether user's total score meets predetermined standard by tracked mobile object mark separately; Tracked mobile object is a kind of foreign matter in electric system power transmission line, and the overall score that meets preassigned shows at least to exist in scape on the scene a foreign matter.
Wherein, described step 2 specifically comprises the following steps:
The first step, the frame of video that deducts first digit vision signal from the frame of video of second digit vision signal obtains difference frame;
Second step, the non-zero region of determining difference frame is mobile object.
Wherein, described step 2 specifically comprises following two steps:
The first step, the weighted mean value that deducts other multiple frame of video of digital video signal from a frame of video of digital video signal obtains difference frame;
Second step, is defined as the mobile object of the non-zero region of difference frame.
Wherein, in described step 4, in digital video signal, locate this mobile object, check and carry out according to the following steps:
The first step, the second difference frame after first difference frame that has mobile object;
Second step, position around, the predetermined field of second difference frame;
The 3rd step, corresponding to the position of second difference frame of first difference frame position that at least has a mobile object.
Wherein, can continue to be positioned at the time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object body, comprise in the time that tracked mobile object persists in digital video signal the nonlinear increase of the mark of this tracked mobile object.
Wherein, predetermined standard is that overall score exceedes predetermined threshold values, shows method that whether overall score meets predetermined standard comprises start at least one vision and audible indicator to user.
Wherein, in described step 1, from digital imaging apparatus Real-time Obtaining digital video signal.
Wherein, in described step 2, can show in real time that difference video signal, this signal comprise from digital video signal to obtain, for helping user to aim at the difference frame of digital imaging apparatus.
Wherein, scene and difference video signal combine in a view.
Wherein, in described step 1, be to obtain digital video signal the digital video file from prerecording.
Wherein, in described step 6, also comprise: the curve map of drawing digital video signal frame number function entirety score.
The invention provides a kind of system of following the trail of mobile object in scene, this system comprises the processor of at least one internal memory that comprises multiple programmed instruction, and described programmed instruction is used for making processor to realize following steps:
Step 1, obtains the digital video signal corresponding with this scene;
Step 2 is identified at least one mobile object in the digital video signal of scene;
Step 3 after at least one mobile object of identification, is found out this mobile object in digital video signal;
Step 4, one period of schedule time after mobile object is identified, continues to be positioned digital video signal;
Step 5, can continue to be positioned at time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object, and the time, longer mark was higher;
Step 6, obtains the overall score of scene in order to show whether user's total score meets predetermined standard by tracked mobile object mark separately; Tracked mobile object is a kind of foreign matter in electric system power transmission line, and the overall score that meets preassigned shows at least to exist in scape on the scene a foreign matter.
Wherein, this system also comprises the digital imaging apparatus that produces digital video signal, and described processor is from digital imaging apparatus Real-time Obtaining digital video signal.
Wherein, this system also comprises display, and processor control display shows the difference frame vision signal for identifying and following the trail of in real time.
Wherein, processor merges scene and difference video signal in view display.
Wherein, when described programmed instruction is used for making processor to realize following steps two, complete in two steps:
The first step, the weighted mean value that deducts other multiple frame of video of digital video signal from a frame of video of digital video signal obtains difference frame;
Second step, is defined as the mobile object of the non-zero region of difference frame.
Wherein, when described programmed instruction is used for making processor to realize following steps two, complete in two steps:
The first step, the weighted mean value that deducts other multiple frame of video of digital video signal from a frame of video of digital video signal obtains difference frame;
Second step, is defined as the mobile object of the non-zero region of difference frame.
Wherein, how long described programmed instruction can continue to be positioned a mark of digital video signal distribution to each tracking mobile object according to following the trail of mobile object after predetermined time; In the time that tracked mobile object persists in digital video signal, the nonlinear increase of the score of tracked mobile object.
The present invention also provides a kind of computer-readable recording medium, comprises the executable multiple programmed instruction of processor, and multiple programmed instruction comprise: first instruction segment that is configured to make to obtain corresponding to scene the processor of digital video signal; Be configured to second instruction segment of the processor that makes the one or more mobile objects in discriminating digit vision signal in scene; Be configured to make, after at least one mobile object of identification, in digital video signal, locate the 3rd instruction segment of the processor of at least one mobile object; After a period of time that is configured to make to be identified at tracked mobile object, persist in digital video signal, follow the trail of the 4th instruction segment of the processor of mobile object; Be configured to make to follow the trail of mobile object to each and after predetermined time, how long can continue to be positioned according to following the trail of mobile object the 5th instruction segment that digital video signal distributes the processor of a mark; Be configured to make to obtain in conjunction with tracking mobile object mark separately the 6th instruction segment of the processor of the overall score of scene; Be configured to make and show whether user's total score meets the 7th instruction segment of the processor of predetermined standard.
Compared to prior art, method and system of following the trail of mobile object in scene provided by the invention, the method is to obtain moving target profile by adjacent two frames in sequence of video images are done to difference frame computing, it can well be applicable to the situation that exists multiple moving targets and video camera to move, in the time there is abnormal object of which movement in monitoring scene, between frame and frame, there will be comparatively significantly difference, two frames subtract each other, obtain the absolute value of two two field picture luminance differences, judge whether it is greater than threshold values and analyzes the kinetic characteristic of video or image sequence, determine and in image, have or not mobile object, can discharge the vegetation of rustling or the noise of of short duration flicker sunlight, and some rudimentary intrinsic noises of digital imaging apparatus, also have the interference of our uninterested mobile object.
Brief description of the drawings
Fig. 1 is the system chart that the embodiment of the present invention is followed the trail of mobile object in scene;
Fig. 2 is the FBD (function block diagram) that the embodiment of the present invention disposes the computing equipment for analyzing corresponding scene digital video signal mobile object Trancking Software;
Fig. 3-6th, the several embodiment of the present invention calculate the schematic diagram of the several different methods of different frame from digital video signal;
Fig. 7 and 8 is schematic diagram of the difference frame of embodiment of the present invention application neighbor filter;
Fig. 9 is the process flow diagram that the embodiment of the present invention is followed the trail of the method for mobile object in scene;
Figure 10 is the process flow diagram that the embodiment of the present invention is identified one or more mobile object methods in digital video signal;
Figure 11 is the process flow diagram that another one embodiment of the present invention identifies one or more mobile object methods in digital video signal;
Embodiment
In order more clearly to explain the present invention, below in conjunction with accompanying drawing, the present invention is further described.
The invention provides a kind of method of following the trail of mobile object in scene, the method comprises that applying the digital video signal obtaining in scene identifies one or more mobile objects.After it is identified, need in digital video signal, locate at least one mobile object.Finally follow the trail of mobile object, one period of schedule time after mobile object is identified, continue to be positioned digital video signal.Can continue to be positioned at the time of digital video signal according to tracked mobile object after by the schedule time, distribute a mark to each tracked mobile object body.Obtain the overall score of scene by tracked mobile object mark separately in order to show whether user's total score meets predetermined standard.
In digital video signal, identification and tracking mobile object are helpful to isolate mobile foreground object from main static background.A technology that realizes this separation is that the difference by calculating between digital video signal frame generates discrepant vision signal.This technology is called as " frame subtracts each other ", and the discrepant vision signal of a frame generation can be called as " frame difference ".The field (intensity is zero or approaches zero) of the different frame of black shows immovable background area from frame to frame.Show in the bright region of different frame may be mobile object or noise (as, the flicker that sunlight is of short duration or the vegetation rustling).
Exceed in the embodiment of threshold values meeting preassigned, threshold values can be according to specific application layer and environmental selection, and user can regulate.For example, a given time, the ad-hoc location that threshold values can use in system adjusts with respect to on-the-spot measurement background mark.Corresponding to interested other mobile objects in tracked mobile object 404 or scene, will be by exceeding " base " background mark score more than suitable selection threshold values, from the background mark of establishing " distinguishing ".
Compared to prior art, the method of following the trail of mobile object in scene the invention provides, by being done to difference frame computing, adjacent two frames in sequence of video images obtain moving target profile, it can well be applicable to the situation that exists multiple moving targets and video camera to move, in the time there is abnormal object of which movement in monitoring scene, between frame and frame, there will be comparatively significantly difference, two frames subtract each other, obtain the absolute value of two two field picture luminance differences, judge whether it is greater than threshold values and analyzes the kinetic characteristic of video or image sequence, determine and in image, have or not mobile object; Can discharge the vegetation of rustling or the noise of of short duration flicker sunlight, and some rudimentary intrinsic noises of digital imaging apparatus, also have the interference of our uninterested mobile object.
The present invention also provides a kind of system of following the trail of mobile object in scene, this system comprises at least one processor and an internal memory that comprises multiple programmed instruction, and the digital video signal that programmed instruction can make processor obtain in scene is identified one or more mobile objects.After it is identified, need in digital video signal, locate at least one mobile object.Finally follow the trail of mobile object, one period of schedule time after mobile object is identified, continue to be positioned digital video signal.Can continue to be positioned at the time of digital video signal according to tracked mobile object after by the schedule time, distribute a mark to each tracked mobile object body.Obtain the overall score of scene by tracked mobile object mark separately in order to show whether user's total score meets predetermined standard.
In a specific embodiment, in scene, follow the trail of the system of mobile object and in digital video signal, identified one or more mobile objects.In order to isolate interested object from noise, system needs a mobile object, once identify this mobile object, continues to observe at least one minimum cycle of digital video signal (for example, a predetermined frame number).In digital video signal, at least observe a minimum cycle for each mobile object, system is followed mobile object and according to by after the minimum period, and it can continue observedly how long to distribute a mark to it in digital video signal.
In this specific embodiment, system combines tracked mobile object mark separately obtains the overall score of scene.System shows to user whether overall score meets predetermined standard.In certain embodiments, the digital video signal of real-time analysis provides a real-time index for whether total score meets standard.
Fig. 1 is the block diagram that the embodiment of the present invention is followed the trail of the system 100 of mobile object in scene, and in the specific embodiment shown in Fig. 1, system 100 comprises the digital imaging apparatus 102 being connected with computing equipment 110 by I/O path 10 6.The communication protocol using at I/O path 10 6 can be any agreement that can process high-speed figure video data.Example include but not limited to, and IEEE 1394(is called as the Apple Computers version of " live wire ") and USB (universal serial bus) (USB) 2.0.
In other embodiment, system 100 does not comprise the proprietary digital imaging apparatus 102 that picture Fig. 1 shows.On the contrary, computing equipment 104 is processed the digital video signal recording, and these signals obtain from an independent digital imaging apparatus, and are loaded in memory device, no matter is inner or outside, and computing equipment 104 can be accessed.The data of digital video capture may be loaded into local connect of memory device by between digital imaging apparatus and memory device, a computer-readable recording medium movably.
An advantage of specific embodiment shown in Fig. 1 is that computing equipment 104 can be analyzed digital video signal, obtains in real time, and whether interested object is successfully followed the trail of at the scene and provided one to show in real time from digital imaging apparatus 102.Following further explanation, the specific embodiment shown in Fig. 1, can provide real-time vision to monitor tracing object, to help user to locate or to change digital imaging apparatus 102.
Although digital imaging apparatus 102 is shown as with computing equipment 104 equipment separating in Fig. 1, in other embodiment, these assemblies of system 100 may be to be integrated in an independently equipment if necessary.
Fig. 2 be the embodiment of the present invention corresponding to scene for analyzing digital video signal, dispose the functional block diagram of the computing equipment 104 of mobile object Trancking Software.For example, computing equipment 104 may be a desktop computer, palm PC, notebook computer, the computer of independently a move-object-tracing system of embedding, or other suitable computing equipments.
In the specific embodiment shown in Fig. 2, processor 202 passes through data bus 204 and input equipment 206, display 208, and internal memory 210, I/O interface 212, memory device 214 communicates.Input equipment 206 may comprise keyboard, mouse or other tip device, or control other input equipments that computing equipment 104 operates.I/O interface 212 may comprise hard disk drive (HDD), optical storage apparatus, flash-memory-based driver, or other memory devices.Internal memory 210 may comprise random access memory (RAM), ROM (read-only memory) (ROM), flash memory.In the specific embodiment shown in Fig. 2, internal memory 210 comprises the video frame buffer 224 that application layer 216 is relevant with, and application layer 216 is configured to Applied Digital image processing techniques in the digital video signal at corresponding scene and follows the trail of mobile object.The digital video framework that application layer 216 is analyzed is cushioned in digital video frame buffer zone 224.
More convenient in order to make to describe in detail, application layer 216 has been divided into one group of functional module: (1) data input/output module 218, (2) analysis engine 220, (3) user interface 222.
Data input/output module 218 can be configured to process input and output data.For example, data input/output module 218 can be managed the numeral-view data from the memory device of digital image apparatus 102 or the next self-contained digital video file of prerecording.
Analysis engine 220 is configured to analyze, the digital video signal at identification and the tracking interested corresponding scene of mobile object in scene.In this course, analysis engine 220 has used multiple digital image processing techniques described below.
The computing equipment 104 of user interface 255 management and user interactions.User interface 222, in conjunction with data input/output module 218, receives the input command (for example, by keyboard and mouse) from user, propagates suitable prompting and user other guidances in computing equipment 104 operating process.
In other specific embodiments, the functional module of above-mentioned application layer 216 can be segmented or the mode of combination except showing in Fig. 2.Meanwhile, the function of the application layer of moving on computing equipment can realize at software firmware, self-defined hardware or any combination or their sub-portfolio.In the specific embodiment shown in Fig. 2, the function of application layer 216 is used to the batch processing instruction executable file of processor 202, is loaded in internal memory 210 before and carries out, and is present in computer-readable recording medium as disk, CD or flash memory etc.For example, the programmed instruction of application layer 216 may be stored in memory device 214.The programmed instruction that forms application layer 216 can be divided into multiple instruction segments or other functional units.
In digital video signal, identification and tracking mobile object are helpful to isolate mobile foreground object from main static background.A technology that realizes this separation is that the difference by calculating between digital video signal frame generates discrepant vision signal.This technology is called as " frame subtracts each other ", and the discrepant vision signal of a frame generation can be called as " frame difference ".The field (intensity is zero or approaches zero) of the different frame of black shows immovable background area from frame to frame.Show in the bright region of different frame may be mobile object or noise (as, the flicker that sunlight is of short duration or the vegetation rustling).
Fig. 3-6th, shows the diagram that several specific embodiments of the present invention calculate the several different methods of different frame from produce the digital video signal of a variant vision signal.
In the specific embodiment showing at Fig. 3, the frame of video 302 that is labeled F1 corresponding to time more early deducts corresponding to the frame of video 302 of late time mark F2 and obtain a difference frame (not showing) in Fig. 3.The technician in this field will appreciate that F1-F2(frame of video afterwards deducts frame of video above) to obtain difference frame be possible.In certain embodiments, the F1 of frame of video 302 and F2 are continuous frame of video.In other some embodiment, may have the frame of video of intervention at the F1 of frame of video 302 and F2.
In the specific embodiment shown in Fig. 4, the F3 that on average deducts frame of video 302 afterwards of the F1 of two early stage frame of video 302 and F2 obtains a difference frame (not being presented in Fig. 4).Again, the relation between frame of video may reverse.Namely, if necessary, after average the substituting of two frame of video 302 deduct early stage frame of video.Specific embodiment just as discussed in Figure 3, the frame of video 302F1 in Fig. 4, F2 and F3 may be continuous frame of video 302, or between them, some intervenes frame of video.
In the specific embodiment shown in Fig. 9, around four frame of video 302(F1 of a frame of video 302F3, F2, F4 and F5) the F3 that on average deducts obtain a difference frame (do not show) in Fig. 9.The viewpoint that this embodiment just reduces speckle and noise in difference frame is desirable especially.
In the embodiment shown in fig. 6, above, the method for Fig. 9 discussion is broad sense.In certain embodiments, weighting factor 304 is applied to frame of video 302F1 around, F2, and F4 and F5 produce weighted mean.
The technician in this field will appreciate that the various embodiment in above-mentioned Fig. 3-6 design the frame of video 302 that cushions varying number in a frame buffer, for example video frame buffer 224.
Once the digital video signal that analysis engine 220 obtains by data input/output module 245 has produced a difference video signal.Mobile object is identified and followed the trail of to analysis engine 220 in digital video signal.As discussing in detail of next step, this difference vision signal that may be known as by application " neighbor filter " completes.
Fig. 7 and Fig. 8 are the figure of the difference frame 402 of embodiment of the present invention application neighbor filter.In Fig. 7, difference frame 402(is labeled as difference frame A) comprise the mobile object 404 being formed by several pixels.Note, in Fig. 7 and 8, black (zero or approach zero) background area is shown as the mobile object 404 that white and relative bright pixel form.In the time that analysis engine 220 is met for the first time one group of pixel relatively becoming clear in difference frame 402 " A ", it will identify this group pixel as mobile object 404.
Fig. 8 is the key diagram that produces late difference frame 402 " B " in Fig. 7 than difference frame 402 " A ".Analysis engine 220 checks around the profile shown in the 408(Fig. 8 of position in this embodiment) neighborhood 406 around, here mobile object 404 is positioned at previous difference frame 402 " A ".
In the example of Fig. 8, analysis engine 220 has been positioned at mobile object 404 within the scope of its new position neighborhood 406.In Fig. 8, the circular boundary of neighborhood 406 is illustrative.Any appropriate border (square, hexagon etc.) of neighborhood 406 all can be used, and depends on specific application layer.The size that the technician in this field be will appreciate that to neighborhood 406 can be adjusted, to optimize the performance for the neighbor filter of specific program.
Technology in Fig. 8 can be called " neighbor filter ", because analysis engine 220 is searched for, at difference frame 402 subsequently, for the part in the mobile object 404 of previous identification afterwards difference frame be near this mobile object 404 in position that previously difference frame 402 is observed for the last time.
When analysis engine 250 has successfully been positioned a specific sufficiently long time of mobile object 404, prove a conclusion, this be not erroneous judgement (as, speckle or noise), mobile object 404 will be followed the trail of and assess to analysis engine 220.For example, in a specific embodiment, mobile object 404 will be followed the trail of and assess to analysis engine 220, after it is positioned at the continuous difference frame of 10 frames 402.For observing a specific mobile object 404 after each difference frame 402, the numerical value score relevant to mobile object 404 increases progressively.Therefore, the time that given mobile object 404 is observed in digital screen signal is longer, and its mark is higher.Such a points-scoring system support is present in relatively long-time mobile object (as the foreign matter of looking for food) slowly at the scene.
Above-mentioned feature can emphasize that tracked mobile object 404 marks are not linear increases in time, and the particular value that score increases progressively and time cycle divide and comprise that above-mentioned ten difference frames 402 are only illustrative.Conventionally in such embodiments,, in the time that mobile object 404 is still present in difference video signal, the advancing the speed of mark of tracked mobile object 404 increases in time.
As mentioned above, the independent score of tracked mobile object 404 can be combined into the overall score of scene.Analysis engine 220 can be configured to according to predetermined standard testing entirety score.In an illustrative embodiment, analysis engine 220 is for more overall score and predetermined threshold values.Predetermined threshold values is exceeded explanation and in scene, has a foreign matter at least.User interface 222 reports whether user meets predetermined standard (as whether exceeded predetermined threshold values).In the status report to user, user interface 222 can be used visual detector, as the text on display 208 or image, or light; Audible indicator is as bleep, the tone, or spoken recording; Or both have both at the same time.
Exceed in the embodiment of threshold values meeting preassigned, threshold values can be according to specific application layer and environmental selection, and user can regulate.For example, a given time, the ad-hoc location that threshold values can use in system adjusts with respect to on-the-spot measurement background mark.Corresponding to interested other mobile objects in tracked mobile object 404 or scene, will be by exceeding " base " background mark score more than suitable selection threshold values, from the background mark of establishing " giving prominence to ".
Fig. 9 is the process flow diagram that the embodiment of the present invention is followed the trail of mobile object method in scene.502, data input/output module 218 is obtained the digital video signal corresponding to this scene.As mentioned above, data input/output module 218 may be obtained a real-time digital video signal from the digital imaging apparatus 102 being connected, and it also can obtain digital video signal from residing in digital video file or other sources that memory device 214 prerecords.504, analysis engine 220 is identified one or more mobile objects 404 in digital video signal.As mentioned above, analysis engine 220 can so do by apply neighbor filter in the difference video signal being made up of difference frame 402.
506, analysis engine 220 is seated and in digital video signal, has a mobile object 404 at least after identification first in 504.508, analysis engine 220 is followed the trail of any one mobile object 404, at least a schedule time after their initial identification, continues to observe in digital video signal.510, how long analysis engine 220 exceeds predetermined time according to tracked mobile object 404 can persist in digital video signal, distributes a mark to each tracked mobile object 404.
512, analysis engine 220 obtains the overall score of scene in conjunction with the mark of each mobile object 404.514, user interface 222 shows to user whether overall score meets predetermined standard.This process is in 516 terminations.
Figure 10 is the process flow diagram that the embodiment of the present invention is identified the method for one or more mobile objects in digital video signal.In Figure 10, the input of process Fig. 9 504 after.602, analysis engine 220 deducts first vision signal 302 from second vision signal 302 and obtains a difference frame 402.604, analysis engine 220 is identified as the non-zero region of the difference frame 402 in mobile object 404, and then this process proceeds to 506 of Fig. 9.
Figure 11 is the process flow diagram that another inventive embodiments is identified one or more mobile object 404 methods in digital video signal.This process Fig. 9 504 after start, 608, analysis engine 220 from first frame of video 302 deduct other multiple frame of video 302 weighted mean produce a difference frame 402.Here, this term " weighted mean " refers to be generally equal special circumstances with the weight that contains average multiple frame of video 302.In addition, 604, analysis engine 220 is identified as the non-zero region of the difference frame 402 in mobile object 404, and then this process proceeds to 506 of Fig. 9.
In one embodiment, computing equipment 104 difference video signal that output is produced by analysis engine 220 approximately in real time depends on the calculating that has the each difference frame 402 of how many frame of video 302 needs to display 208(, postpone because the buffering of frame of video 302 difference video signal relevant to digital video signal has some).Show that described difference video signal can aim at or reorientate digital imaging apparatus 102 and offer help to user.In order further to help user aim at or reorientate digital imaging apparatus 102, in a view, the low intensive version of whole scene can combine with difference video signal.For example, what each pixel of the digital video signal of corresponding scene can be determined in advance be greater than, and individual nonzero value is divided, and the view of this scene that has reduced intensity and difference video signal can be applied on display 208.
Disclosed is above only several specific embodiment of the present invention, but the present invention is not limited thereto, and the changes that any person skilled in the art can think of all should fall into protection scope of the present invention.
Claims (10)
1. a method of following the trail of mobile object in scene, is characterized in that, comprises following step:
Step 1, obtains the digital video signal corresponding with this scene;
Step 2 is identified at least one mobile object in the digital video signal of scene;
Step 3 after at least one mobile object of identification, is found out this mobile object in digital video signal;
Step 4, one period of schedule time after mobile object is identified, continues to be positioned digital video signal;
Step 5, can continue to be positioned at time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object, and the time, longer mark was higher;
Step 6, obtains the overall score of scene in order to show whether user's total score meets predetermined standard by tracked mobile object mark separately; Tracked mobile object is a kind of foreign matter in electric system power transmission line, and the overall score that meets preassigned shows at least to exist in scape on the scene a foreign matter.
2. method of following the trail of mobile object in scene according to claim 1, is characterized in that, described step 2 specifically comprises the following steps:
The first step, the frame of video that deducts first digit vision signal from the frame of video of second digit vision signal obtains difference frame;
Second step, the non-zero region of determining difference frame is mobile object.
3. method of following the trail of mobile object in scene according to claim 1, is characterized in that, described step 2 specifically comprises following two steps:
The first step, the weighted mean value that deducts other multiple frame of video of digital video signal from a frame of video of digital video signal obtains difference frame;
Second step, is defined as the mobile object of the non-zero region of difference frame.
4. according to the method for following the trail of mobile object in scene described in claim 2 or 3, it is characterized in that, in described step 4, in digital video signal, locate this mobile object, check and carry out according to the following steps:
The first step, the second difference frame after first difference frame that has mobile object;
Second step, position around, the predetermined field of second difference frame;
The 3rd step, corresponding to the position of second difference frame of first difference frame position that at least has a mobile object.
5. method of following the trail of mobile object in scene according to claim 1, it is characterized in that, can continue to be positioned at the time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object body, comprise in the time that tracked mobile object persists in digital video signal the nonlinear increase of the mark of this tracked mobile object.
6. method of following the trail of mobile object in scene according to claim 2, is characterized in that, in described step 1, from digital imaging apparatus Real-time Obtaining digital video signal; In described step 2, can show in real time that difference video signal, this signal comprise from digital video signal to obtain, for helping user to aim at the difference frame of digital imaging apparatus.
7. the invention provides a kind of system of following the trail of mobile object in scene, it is characterized in that, this system comprises the processor of at least one internal memory that comprises multiple programmed instruction, and described programmed instruction is used for making processor to realize following steps:
Step 1, obtains the digital video signal corresponding with this scene;
Step 2 is identified at least one mobile object in the digital video signal of scene;
Step 3 after at least one mobile object of identification, is found out this mobile object in digital video signal;
Step 4, one period of schedule time after mobile object is identified, continues to be positioned digital video signal;
Step 5, can continue to be positioned at time of digital video signal after by the schedule time according to tracked mobile object, distribute a mark to each tracked mobile object, and the time, longer mark was higher;
Step 6, obtains the overall score of scene in order to show whether user's total score meets predetermined standard by tracked mobile object mark separately; Tracked mobile object is a kind of foreign matter in electric system power transmission line, and the overall score that meets preassigned shows at least to exist in scape on the scene a foreign matter.
8. system of following the trail of mobile object in scene according to claim 7, is characterized in that, this system also comprises the digital imaging apparatus that produces digital video signal, and described processor is from digital imaging apparatus Real-time Obtaining digital video signal.
9. system of following the trail of mobile object in scene according to claim 7, is characterized in that, this system also comprises display, and processor control display shows the difference frame vision signal for identifying and following the trail of in real time; Processor merges scene and difference video signal in view display.
10. system of following the trail of mobile object in scene according to claim 7, it is characterized in that, how long described programmed instruction can continue to be positioned a mark of digital video signal distribution according to following the trail of mobile object after predetermined time if being followed the trail of mobile object to each; In the time that tracked mobile object persists in digital video signal, the nonlinear increase of the score of tracked mobile object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410274348.XA CN104021576A (en) | 2014-06-18 | 2014-06-18 | Method and system for tracking moving objects in scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410274348.XA CN104021576A (en) | 2014-06-18 | 2014-06-18 | Method and system for tracking moving objects in scene |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104021576A true CN104021576A (en) | 2014-09-03 |
Family
ID=51438313
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410274348.XA Pending CN104021576A (en) | 2014-06-18 | 2014-06-18 | Method and system for tracking moving objects in scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104021576A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678730A (en) * | 2014-11-17 | 2016-06-15 | 西安三茗科技有限责任公司 | Camera movement self-detecting method on the basis of image identification |
| CN105825619A (en) * | 2016-03-25 | 2016-08-03 | 南京第五十五所技术开发有限公司 | High-tension cable floater alarm method based on image processing |
| CN106331462A (en) * | 2015-06-25 | 2017-01-11 | 宇龙计算机通信科技(深圳)有限公司 | Method and device for shooting track pictures, as well as mobile terminal |
| CN109272529A (en) * | 2017-07-18 | 2019-01-25 | 信浓绢糸株式会社 | Video analysis method |
| CN109643452A (en) * | 2016-08-12 | 2019-04-16 | 高通股份有限公司 | The method and system of lost objects tracker is maintained in video analysis |
-
2014
- 2014-06-18 CN CN201410274348.XA patent/CN104021576A/en active Pending
Non-Patent Citations (3)
| Title |
|---|
| X. RONG LI, YAAKOV BAR-SHALOM: "Tracking in clutter with nearest neighbor filters: analysis and performance", 《IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS》 * |
| 屈晶晶，辛云宏: "连续帧间差分与背景差分相融合的运动目标检测方法", 《光子学报》 * |
| 李然 等: "基于视频监视系统的运动目标检测技术的实现", 《电力系统通信》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678730A (en) * | 2014-11-17 | 2016-06-15 | 西安三茗科技有限责任公司 | Camera movement self-detecting method on the basis of image identification |
| CN106331462A (en) * | 2015-06-25 | 2017-01-11 | 宇龙计算机通信科技(深圳)有限公司 | Method and device for shooting track pictures, as well as mobile terminal |
| CN105825619A (en) * | 2016-03-25 | 2016-08-03 | 南京第五十五所技术开发有限公司 | High-tension cable floater alarm method based on image processing |
| CN109643452A (en) * | 2016-08-12 | 2019-04-16 | 高通股份有限公司 | The method and system of lost objects tracker is maintained in video analysis |
| CN109272529A (en) * | 2017-07-18 | 2019-01-25 | 信浓绢糸株式会社 | Video analysis method |
| CN109272529B (en) * | 2017-07-18 | 2023-07-14 | 信浓绢糸株式会社 | Video analysis method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104021576A (en) | Method and system for tracking moving objects in scene | |
| CN108399359B (en) | Real-time fire detection early warning method under video sequence | |
| CN111489342B (en) | Video-based flame detection method, system and readable storage medium | |
| CN101231696A (en) | Method and system for detection of hangover | |
| EP3282387A1 (en) | Fire detection method, fire detection apparatus and electronic equipment | |
| CN112257632A (en) | Transformer substation monitoring system based on edge calculation | |
| CN104346626A (en) | SF6 gas leakage online mode recognition method based on image recognition | |
| JP2010015469A (en) | Still area detection method, and apparatus, program and recording medium therefor | |
| CN111860324A (en) | High-frequency component detection and color identification fire early warning method based on wavelet transformation | |
| CN108184098B (en) | Method and system for monitoring safety area | |
| KR101044903B1 (en) | Fire detecting method using hidden markov models in video surveillance and monitoring system | |
| US20100310120A1 (en) | Method and system for tracking moving objects in a scene | |
| CN107729811B (en) | Night flame detection method based on scene modeling | |
| US20190325587A1 (en) | Gas detection-use image processing device, gas detection-use image processing method, and gas detection-use image processing program | |
| CN113902694A (en) | Target detection method based on dynamic and static combination | |
| JPH09293141A (en) | Mobile object detection device | |
| CN103425958A (en) | Method for detecting non-movable objects in video | |
| JPH06308256A (en) | Cloudy fog detecting method | |
| CN111541877A (en) | Automatic monitoring system for substation equipment | |
| CN113221603A (en) | Method and device for detecting shielding of monitoring equipment by foreign matters | |
| Riyadi et al. | Vision based flame detection system for surveillance camera | |
| KR20120121627A (en) | Object detection and tracking apparatus and method thereof, and intelligent surveillance vision system using the same | |
| KR102516172B1 (en) | System, Method for Extracting Color of Foreground and Computer Readable Record Medium Thereof | |
| CN104820818A (en) | Fast recognition method for moving object | |
| JP6618438B2 (en) | Foreground region extraction apparatus, method, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140903 |