CN102136147B - Target detecting and tracking method, system and video monitoring device - Google Patents
Target detecting and tracking method, system and video monitoring device Download PDFInfo
- Publication number
- CN102136147B CN102136147B CN2011100694470A CN201110069447A CN102136147B CN 102136147 B CN102136147 B CN 102136147B CN 2011100694470 A CN2011100694470 A CN 2011100694470A CN 201110069447 A CN201110069447 A CN 201110069447A CN 102136147 B CN102136147 B CN 102136147B
- Authority
- CN
- China
- Prior art keywords
- target
- pixel
- image
- tracking
- ratio
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention is applicable to the technical field of intelligent information processing, and provides a target detecting and tracking method, system and video monitoring device. The method comprises the steps as follows: step 1, the video data input from a client is received; any image in the video data is collected, and a pixel threshold value is set; step 2, a specific value of a target pixel point number in an mage to an overall pixel point number in the image is obtained according to the pixel threshold value; then whether the specific value is within a preset scope or not is judged; if not, the step 3 is actuated; if so, then step 4 is actuated; step 3, the pixel threshold value is adjusted to return to step 2; and step 4, the property information of the pixel points of the target in the image is obtained according to the pixel threshold value; and the tracked target is obtained according to the property information of the pixel points. The invention solves the problems that a video monitoring system can not effectively and accurately detect the target, enhances the anti-jamming capacity for target detecting, quickens the speed for target obtaining, and further improves the accuracy and position precision of the target.
Description
Technical field
The invention belongs to the intelligent information processing technology field, relate in particular to a kind of target detection and tracking, system and video monitoring equipment.
Background technology
In today of supervisory system fast development, system scale is increasing, and the control point is more and more, and the staff also has no time to attend in the face of increasing image.The intelligent monitoring product that thereupon expedites the emergence of out will have been played the part of the role of intellectual analysis, be the trend that the intellectuality of representative becomes the front end production development with the mobile tracking.Since the mobile tracking quick was released, more and more ball machine manufacturers that suffer from the serious present situation of ball machine function homogeneity were decided to be the important directions that the ball machine develops to mobile tracking.Up to the present, the intelligent-tracking clipping the ball has experienced the development of two generations.The first generation is based on motion detection in the zone, usually causes tracking target to be switched when a plurality of moving target is arranged in the zone.Second on behalf of locking tracking, and target is not easy to switch, but present product still has a lot of defectives and deficiency.
In the target detection link, general target and background separation strategy have two kinds.A kind of is that the frame of fixed threshold is poor; Be provided with in threshold value and can obtain sport foreground fast with less operand when reasonable; But threshold value is provided with when excessive or too small, and it is very few or too much prospect to occur, and ambient lighting is when changing; Threshold value can not change and change along with ambient lighting, causes the omission or the flase drop of prospect.A kind of is the Gauss model that pursues the pixel adaptive threshold; But the Gauss model operand is big; And need the study of accumulation multiframe just can detect prospect preferably, the static long period of video camera just can obtain the stabilizing effective foreground target, and video camera is static to next time detection or correction target from moving; And when being applied to the axially-movable video camera, the gross error that static background becomes prospect can appear.
In the target following link, effect is preferably the signature tracking based on field color, but method in common is to do the target area as a whole, extracts color histogram, matees at the corresponding close position of next frame image.Histogram is the statistic in the zone, does not comprise positional information.Yet actual target is not necessarily monochromatic, people especially, and upper garment and trousers color often have than big-difference; If lump together statistics to these two colors, effect and jacket and trousers color put upside down or the target class of other color seemingly, produce mistake and mate; Target is switched easily; Be that the color of object piece does not have spatial information, cause following the tracks of failure, cause the apparent error on the visual effect.Be exactly in addition target exist a kind of during than the large tracts of land color region similar with background tracing area can go to background and get on, cause following the tracks of failure.Moreover; Piecemeal marginal position in target area is provided with also has tangible influence to tracking effect; Can the separatrix be set in the apparent in view position of color contrast when manually following the tracks of, yet, from motion tracking the time; Nobody's intervention is not demarcated or separatrix that the fixed position is set is not to select preferably.Though some ready-made targets are arranged by the color block division method, all operand is big, and need does further processed and could make things convenient for use.
Summary of the invention
The purpose of the embodiment of the invention is to provide a kind of target detection and tracking, system and video monitoring equipment; Be intended to solve since existing tracking when detecting target; Receive the target influence of many factors such as background or environment dynamic change on every side, cause supervisory system can not realize effectively and the exactly problem of tracking target.
The embodiment of the invention is achieved in that a kind of object detection method, and said method comprises the steps:
Step 1 receives the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
Step 2 is obtained the ratio of pixel number total in pixel number and the said image of target in the said image according to said pixel threshold, is specially the absolute value that obtains said image and preset background image respective point pixel intensity value difference; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value, and the ratio of total pixel number in the pixel number of obtaining target in the said image and the said image continues to judge that said ratio is whether in the scope of presetting; Not; Then execution in step 3, are, then execution in step 4;
Step 3 is adjusted said pixel threshold, and returns step 2;
Step 4 is obtained the pixel attribute information of target in the said image according to said pixel threshold, obtains the target of tracking according to the pixel attribute information of said target.
Another purpose of the embodiment of the invention is to provide a kind of method that adopts above-mentioned object detection method to realize target following, and said method comprises the steps:
The target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of said tracking target area piece for preset number;
Judge the target that whether has said tracking in the image of subsequent frame according to the characteristic information of said target area piece, the output tracking result.
Another purpose of the embodiment of the invention is to provide a kind of object detection system, and said system comprises video data input block, ratio acquiring unit, ratio in judgement unit, threshold adjustment unit and object detection unit, wherein:
The video data input block is used to receive the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
The ratio acquiring unit; Be used for pixel threshold according to the setting of video data input block; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image; Be specially the absolute value of the background image respective point pixel intensity value difference that obtains said image and preset; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image, perhaps in the output result of said ratio in judgement unit for not the time, obtain the ratio of pixel number total in pixel number and the said image of target in the said image according to the pixel threshold of said threshold adjustment unit output;
The ratio in judgement unit, whether the ratio that is used to judge the output of ratio acquiring unit is in preset scope;
Threshold adjustment unit is used in the output result of ratio in judgement unit for not the time adjustment pixel threshold; And
Object detection unit is used in the output result of ratio in judgement unit when being, obtains the pixel attribute information of target in the image according to pixel threshold, obtains the target of tracking according to said pixel attribute information.
Another purpose of the embodiment of the invention is to provide a kind of Target Tracking System that comprises above-mentioned object detection system, and said system also comprises:
Target area piece acquiring unit is used for the target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of the said tracking target area piece for preset number; And
The tracking results output unit is used for judging according to the characteristic information of said target area piece whether the image of subsequent frame exists the target of said tracking, output tracking result.
Another purpose of the embodiment of the invention is to provide a kind of video monitoring equipment that comprises above-mentioned object detection system or Target Tracking System.
The embodiment of the invention is through obtaining in the video data in any two field picture the ratio of total pixel number in the pixel number of target and this image according to pixel threshold; Judge that this ratio whether in preset scope, not, then adjusts this pixel threshold; And obtain the ratio of pixel number total in pixel number and this image of target in this image again according to adjusted pixel threshold; Abovementioned steps is carried out in circulation, in preset scope, then obtains the pixel attribute information of target in this image until this ratio according to this pixel threshold; Obtain the target of tracking according to the pixel attribute information of this target; Solved when automatically track target,, caused video monitoring equipment can not realize effectively and exactly detecting the problem of target owing to receive the target influence of many factors such as background or environment dynamic change on every side; Make the antijamming capability of target detection strengthen; Obtain speeding up of target, the confidence level and the positional precision of target are higher, and video monitoring equipment is intelligence, reliable more.
Description of drawings
Fig. 1 is the realization flow figure of a kind of object detection method of providing of first embodiment of the invention;
Fig. 2 is the realization flow figure of a kind of object detection method of providing of second embodiment of the invention;
Fig. 3 is the realization flow figure of a kind of method for tracking target of providing of third embodiment of the invention;
Fig. 4 is the structural drawing of a kind of object detection system of providing of fourth embodiment of the invention;
Fig. 5 is the structural drawing of a kind of Target Tracking System of providing of fifth embodiment of the invention.
Fig. 6 is the concrete structure figure of target area piece acquiring unit 56 in a kind of Target Tracking System of providing of fifth embodiment of the invention.
Embodiment
In order to make the object of the invention, technical scheme and advantage clearer,, the present invention is further elaborated below in conjunction with accompanying drawing and embodiment.Should be appreciated that specific embodiment described herein only in order to explanation the present invention, and be not used in qualification the present invention.
The embodiment of the invention is through obtaining in the video data in any two field picture the ratio of total pixel number in the pixel number of target and this image according to pixel threshold, judge that this ratio is whether in preset scope, not; Then adjust this pixel threshold; And obtain the ratio of pixel number total in pixel number and this image of target in this image again according to adjusted pixel threshold, abovementioned steps is carried out in circulation, until this ratio in preset scope; Then obtain the pixel attribute information of target in this image according to this pixel threshold; Obtain the target of tracking according to the pixel attribute information of this target, thereby make the antijamming capability of target detection strengthen, obtain speeding up of target; The confidence level and the positional precision of target are higher, and video monitoring system is intelligence, reliable more.
The embodiment of the invention provides a kind of object detection method, and said method comprises the steps:
Step 1 receives the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
Step 2 is obtained the ratio of pixel number total in pixel number and the said image of target in the said image according to said pixel threshold, is specially the absolute value that obtains said image and preset background image respective point pixel intensity value difference; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value, and the ratio of total pixel number in the pixel number of obtaining target in the said image and the said image continues to judge that said ratio is whether in the scope of presetting; Not; Then execution in step 3, are, then execution in step 4;
Step 3 is adjusted said pixel threshold, and returns step 2;
Step 4 is obtained the pixel attribute information of target in the said image according to said pixel threshold, obtains the target of tracking according to the pixel attribute information of said target.
The embodiment of the invention also provides a kind of method for tracking target that adopts above-mentioned object detection method, and said method also comprises the steps:
The target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of said tracking target area piece for preset number;
Judge the target that whether has said tracking in the image of subsequent frame according to the characteristic information of said target area piece, the output tracking result.
The embodiment of the invention also provides a kind of object detection system, and said system comprises video data input block, ratio acquiring unit, ratio in judgement unit, threshold adjustment unit and object detection unit, wherein:
The video data input block is used to receive the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
The ratio acquiring unit; Be used for pixel threshold according to the setting of video data input block; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image; Be specially the absolute value of the background image respective point pixel intensity value difference that obtains said image and preset; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image, perhaps in the output result of said ratio in judgement unit for not the time, obtain the ratio of pixel number total in pixel number and the said image of target in the said image according to the pixel threshold of said threshold adjustment unit output;
The ratio in judgement unit, whether the ratio that is used to judge the output of ratio acquiring unit is in preset scope;
Threshold adjustment unit is used in the output result of ratio in judgement unit for not the time adjustment pixel threshold; And
Object detection unit is used in the output result of ratio in judgement unit when being, obtains the pixel attribute information of target in the image according to pixel threshold, obtains the target of tracking according to said pixel attribute information.
The embodiment of the invention also provides a kind of above-mentioned object detection system Target Tracking System that comprises, said system also comprises:
Target area piece acquiring unit is used for the target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of the said tracking target area piece for preset number; And
The tracking results output unit is used for judging according to the characteristic information of said target area piece whether the image of subsequent frame exists the target of said tracking, output tracking result.
The embodiment of the invention also provides a kind of video monitoring equipment that comprises above-mentioned object detection system or Target Tracking System.
The embodiment of the invention is through obtaining in the video data in any two field picture the ratio of total pixel number in the pixel number of target and this image according to pixel threshold; Judge that this ratio whether in preset scope, not, then adjusts this pixel threshold; And obtain the ratio of pixel number total in pixel number and this image of target in this image again according to adjusted pixel threshold; Abovementioned steps is carried out in circulation, in preset scope, then obtains the pixel attribute information of target in this image until the ratio that obtains according to this pixel threshold; Obtain the target of tracking according to the pixel attribute information of this target; Solved when automatically track target,, caused video monitoring system can not realize effectively and exactly detecting the problem of target owing to receive the target influence of many factors such as background or environment dynamic change on every side; Make the antijamming capability of target detection strengthen; Obtain speeding up of target, the confidence level and the positional precision of target are higher, and video monitoring system is intelligence, reliable more.
Below in conjunction with specific embodiment concrete realization of the present invention is described in detail:
Embodiment one:
Fig. 1 shows the realization flow of a kind of object detection method that first embodiment of the invention provides, and details are as follows:
In step S101, receive the video data of client input, gather any two field picture in this video data, and pixel threshold is set.
In embodiments of the present invention; The video data of client input is except comprising any two field picture; The maximum frame number that allows when also comprising the initial alert regional extent on video pictures that is provided with in advance, the monitor event information of following the tracks of triggering, object matching, preset background image etc., pixel threshold is by the experimental analysis gained.
In step S102, obtain the ratio of pixel number total in pixel number and this image of target in this image according to this pixel threshold, judge this ratio whether in preset scope, not, execution in step S103 then is, then execution in step S104.
Wherein, preset scope is the reasonable distributed area of ratio of total pixel number of pixel number and its place image of target in the normal monitor video scene that obtains according to historical experimental analysis.
Wherein, the ratio that obtains pixel number total in pixel number and this image of target in this image according to this pixel threshold is specially:
Obtain the absolute value of this image and preset background image respective point pixel intensity value difference, this preset background image is not comprising of obtaining in advance of the image during target to be detected;
Pixel in corresponding this image is confirmed as the pixel of target during greater than this pixel threshold with absolute value, obtains the ratio of pixel number total in pixel number and this image of target in this image.
In step S103, adjust this pixel threshold, and return step S102.
In embodiments of the present invention, the step of the pixel threshold that adjustment is preset is: if in this image in the pixel number of target and this image ratio of total pixel number then the pixel threshold of presetting is added 1 greater than preset maximum magnitude value; If this ratio less than preset minimum zone value, then subtracts 1 with preset pixel threshold, upgrade preset pixel threshold.In the practical implementation process; When video monitoring equipment moves; If compare the pixel brightness value of this image with the background image respective point of presetting; Then can cause lot of background to be identified as prospect by mistake, make that the pixel that obtains target is inaccurate, need not upgrade preset pixel threshold this moment; When video monitoring equipment is motionless, add up the ratio of pixel number total in pixel number and this image of target in this image, according to step S102, judge whether to upgrade preset threshold value.
In step S104, obtain the pixel attribute information of target in this image according to this pixel threshold, obtain the target of tracking according to the pixel attribute information of this target.
Wherein, the pixel attribute information includes but not limited to information such as pixel brightness value, pixel position.
In embodiments of the present invention, this object detection method has realized obtaining along with time and Changes in weather adaptive change the parameter of target, makes to obtain speeding up of target to be detected, and the ambient light of target detection adapts to occasion greatly to be expanded.
Embodiment two:
Fig. 2 shows the realization flow figure of a kind of object detection method that second embodiment of the invention provides, and details are as follows:
In embodiments of the present invention, the particular content of step S201, step S202, step S203 is corresponding one by one respectively at the particular content of step S101, step S102, step S103 among the embodiment one.
In step S204, obtain the pixel attribute information of target in this image according to this pixel threshold, the pixel attribute information according to this target obtains alternative tracking target.
In step S205, the alternative tracking target characteristic of obtaining and the characteristic of storage are in advance mated, judge whether to exist the alternative tracking target of coupling, be execution in step S206 then, otherwise execution in step S207.
Wherein, The characteristic of storage can also can be the zone that is provided with in advance, for example for characteristics such as the shape that is stored in people in the database, automobile etc. in advance, Aspect Ratio, area sizes in advance; The alternative target that appears in the zone that is provided with in advance then is the target that will follow the tracks of; Automatically delete other alternative target, also can prepare the target of tracking through manual selection in addition, perhaps manually selection combines with automatic the selection.In the practical implementation process; When having the target of following the tracks of, the pixel of target concentrates on motion target area in a large number in this sequence image that obtains, and handles the back through morphology and forms effective connected region; Be alternative tracking target, more than one of at this moment alternative tracking target possibility.And when the target of following the tracks of does not exist; Pixel positional information according to target in this sequence image that obtains; Learn that this pixel is that discrete and random distributes; The point that these Discrete Stochastic distribute can not form effective connected region after handling through two-value gray scale morphology, does not promptly have alternative tracking target.
In step S206, the alternative target of output coupling is as the target of following the tracks of.
In step S207, judge whether the number of times of images acquired from video data surpasses preset number of times, be execution in step S208 then, otherwise execution in step S209.
In step S208, withdraw from trace routine, this moment, supervisory system can be introduced the video data of client input again.
In the practical implementation process; When not having the alternative tracking target of coupling; Also can store the alternative target of being obtained according to step S205, if surpass preset number of times when the number of times of images acquired from video data, when still not having the alternative tracking target of coupling; From the alternative target of being stored, the alternative target of choosing the most approaching coupling target is as tracking target.
In step S209, the next frame image of any two field picture in the video data of collection client input returns execution in step S202.
In embodiments of the present invention, the sensitivity that supervisory system can keep higher all the time automatically detects target, and has rapid reaction and the little characteristics of calculated amount.
Embodiment three:
Fig. 3 shows the realization flow figure of a kind of method for tracking target that third embodiment of the invention provides, and details are as follows:
In step S301, receive the video data of client input, gather any two field picture in this video data, and pixel threshold is set.
In step S302, obtain the ratio of pixel number total in pixel number and this image of target in this image according to this pixel threshold, judge this ratio whether in preset scope, not, execution in step S303 then is, then execution in step S304.
In step S303, this pixel threshold of root adjustment, execution in step S302.
In step S304, obtain the pixel attribute information of target in this image according to this pixel threshold, obtain the target of tracking according to the pixel attribute information of this target.
In step S305, the target of this tracking is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of this tracking target that the region unit of preset number forms of serving as reasons.
In embodiments of the present invention, step S305 concrete steps are:
The target of this tracking is divided into the identical object block of size according to certain orientation, obtains every color histogram;
According to this color histogram, the color similarity degree between more adjacent two object block;
According to similarity size, with the target identification of the following the tracks of target that the region unit of preset number forms of serving as reasons.
In the practical implementation process, the target of this tracking is divided into the identical object block of size according to certain orientation can be for being divided into the piecemeal of identical size to target according to vertical or horizontal direction, for example; Be divided into four of identical size according to vertical direction; Calculate the color histogram of each object block, histogrammic similarity between more adjacent two then obtains 3 similarities; Color differs less between big two of similarity, and two colors that similarity is little differ bigger.According to the magnitude relationship between 3 similarities, merge the maximum object block of similarity, reconfiguring former object block becomes new region unit, and the number of new region piece preestablishes.Suppose that predefined region unit number is 2, at first merging two maximum object block of similarity is a region unit.Then, if the similarity of middle 2 object block is maximum, then former whole object is second region unit; If the similarity of middle 2 object block is minimum, then merging other 2 object block is second region unit; Otherwise, delete the object block that is in marginal position in 2 minimum object block of similarity, remaining object block is second region unit.
In step S306, judge the target that whether has this tracking in the image of subsequent frame according to the characteristic information of this target area piece, the output tracking result.
Wherein the characteristic information of target area piece can but be not limited to utilize the relative position information between color, shape, texture and the target area piece of target area piece to represent.In the practical implementation process; After obtaining the target that the characteristic information that utilizes the target area piece representes; In the image of subsequent frame, search for the zone of the preset range adjacent, judge the target that whether has the tracking of coupling in the follow-up two field picture that searches with this target location.The tracking results of output comprises whether having the information of mating target, the positional information of coupling target etc.
In embodiments of the present invention; To be divided into target according to image similarity by the target of single area sign, and not only make target stronger with the discrimination of background or jamming target on every side by the higher region unit sign of a plurality of discriminations; Also embody the relative position information between a plurality of piecemeals in the target; Make the antijamming capability of target following stronger, coupling has higher success rate, and has improved the accuracy rate and the precision of following the tracks of greatly.
One of ordinary skill in the art will appreciate that all or part of step that realizes in the foregoing description method is to instruct relevant hardware to accomplish through program; Described program can be stored in the computer read/write memory medium; Described storage medium is like ROM/RAM, disk, CD etc.
Embodiment four:
Fig. 4 shows the structure of a kind of object detection system that fourth embodiment of the invention provides, and for the ease of explanation, only shows the part relevant with the embodiment of the invention.This object detection system comprises video data input block, ratio acquiring unit, ratio in judgement unit, threshold adjustment unit and object detection unit.
Video data input block 41 receives the video data of client input, gathers any two field picture in this video data, and pixel threshold is set.
The pixel threshold that ratio acquiring unit 42 is provided with according to the video data input block; Obtain the ratio of pixel number total in pixel number and this image of target in this image; Perhaps in the output result of this ratio in judgement unit for not the time, obtain the ratio of pixel number total in pixel number and this image of target in this image according to the pixel threshold of this threshold adjustment unit output.
Ratio in judgement unit 43 judges that the ratio of ratio acquiring unit output is whether in preset scope.
In embodiments of the present invention, threshold adjustment unit 44 is specially:
For not the time, when the ratio of the pixel number of target in this image and all pixel numbers during greater than the maximum magnitude value preset, preset pixel threshold adds 1 in the output result of ratio in judgement unit; When the ratio of the pixel number of target in this image and all pixel numbers during, then preset pixel threshold is subtracted 1 less than preset minimum zone value.
In embodiments of the present invention, object detection unit 45 is specially:
Objective attribute target attribute acquiring unit 451 is used in the output result of ratio in judgement unit obtaining the pixel attribute information of target in this image according to this pixel threshold when being;
Alternative target acquiring unit 452 is used for the pixel attribute information according to this target, obtains alternative tracking target; And
Tracking target acquiring unit 453 is used for the characteristic information of the alternative tracking target characteristic information that obtains with storage in advance mated, and obtains the target of tracking.
In embodiments of the present invention, this system can be used for video monitoring equipment.
Embodiment five:
Fig. 5 shows the structure of a kind of Target Tracking System that fifth embodiment of the invention provides, and for the ease of explanation, only shows the part relevant with the embodiment of the invention.
Video data input block 51 receives the video data of client input, gathers any two field picture in this video data, and pixel threshold is set.
The pixel threshold that ratio acquiring unit 52 is provided with according to video data input block 51; Obtain the ratio of pixel number total in pixel number and this image of target in this image; Perhaps in the output result of ratio in judgement unit 53 for not the time, obtain the ratio of pixel number total in pixel number and this image of target in this image according to the pixel threshold of these threshold adjustment unit 54 outputs.
Ratio in judgement unit 53 judges that the ratio of ratio acquiring unit 52 output is whether in preset scope.
Target area piece acquiring unit 56 is divided into the identical object block of size with the target of the tracking of object detection unit 55 outputs according to certain orientation; According to the color similarity degree magnitude relationship between the object block, with the target identification of this tracking target that the region unit of preset number forms of serving as reasons.
In embodiments of the present invention, target area piece acquiring unit 56 comprises that specifically target initially divides module unit 561, similarity acquiring unit 562 and target identification unit 563, and is as shown in Figure 6, wherein:
Target is initially divided module unit 561, is used for the target of this tracking is divided into the identical object block of size according to certain orientation, obtains every color histogram;
Similarity acquiring unit 562 is used for according to this color histogram, the color similarity degree between more adjacent two object block; And
Target identification unit 563 is used for according to the similarity magnitude relationship, with the target identification of the following the tracks of target that the region unit of preset number forms of serving as reasons.
Tracking results output unit 57 is judged the target that whether has this tracking in the image of subsequent frame, output tracking result according to the characteristic information of target area piece.
Tracking results processing unit 58 judges according to the tracking results of tracking results output unit 57 outputs whether preset monitor event takes place, and is, then the behavior of target is made a response, and not, then continues tracking target.
In the practical implementation process, the reposition after target travel departs from the distance value that the picture center surpass to be set, for example during 1/4th picture size; Perhaps target exceeds predefined position range; Perhaps in the preset time, do not find the target of coupling etc., think that then preset monitor event takes place, and sends control information to supervisory system; The control watch-dog is made a response; Such as, rotate watch-dog and make target remain at the picture center, perhaps watch-dog returns to initial position and follows the tracks of again etc.
In embodiments of the present invention, this system can be used for video monitoring equipment.
The embodiment of the invention is through obtaining in the video data in any two field picture the ratio of total pixel number in the pixel number of target and this image according to pixel threshold, judge that this ratio is whether in preset scope, not; Then adjust this pixel threshold; And obtain the ratio of pixel number total in pixel number and this image of target in this image again according to adjusted pixel threshold, abovementioned steps is carried out in circulation, until this ratio in preset scope; Then obtain the pixel attribute information of target in this image according to this pixel threshold; Obtain the target of tracking according to the pixel attribute information of this target, solved, cause the antijamming capability of video monitoring system target detection relatively poor owing to receive the influence of many factors such as background or environment dynamic change on every side; Can not effectively and exactly detect the problem of target; Make the antijamming capability of target detection strengthen, obtain speeding up of target, the accuracy and the positional precision of target are improved.
The above is merely preferred embodiment of the present invention, not in order to restriction the present invention, all any modifications of within spirit of the present invention and principle, being done, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. an object detection method is characterized in that, said method comprises the steps:
Step 1 receives the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
Step 2 is obtained the ratio of pixel number total in pixel number and the said image of target in the said image according to said pixel threshold, is specially the absolute value that obtains said image and preset background image respective point pixel intensity value difference; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value, and the ratio of total pixel number in the pixel number of obtaining target in the said image and the said image continues to judge that said ratio is whether in the scope of presetting; Not; Then execution in step 3, are, then execution in step 4;
Step 3 is adjusted said pixel threshold, and returns step 2;
Step 4 is obtained the pixel attribute information of target in the said image according to said pixel threshold, obtains the target of tracking according to the pixel attribute information of said target.
2. the method for claim 1 is characterized in that, the step of the said pixel threshold of said adjustment specifically comprises:
When the ratio of pixel number total in the pixel number of target in the said image and the said image during, said pixel threshold is added 1 greater than preset maximum magnitude value;
When said ratio during, said pixel threshold is subtracted 1 less than preset minimum zone value.
3. the method for claim 1 is characterized in that, said pixel attribute information comprises brightness value, the positional information of pixel.
4. the method for claim 1 is characterized in that, said step 4 specifically comprises:
Obtain the pixel attribute information of target in the said image according to said pixel threshold, the pixel attribute information according to said target obtains alternative tracking target;
The alternative tracking target characteristic of obtaining and the characteristic of storage are in advance mated, when having the alternative tracking target of coupling, then confirm the target of target for following the tracks of of this coupling;
When not having the alternative tracking target of coupling; Then gather the next frame image of any two field picture in the said video data; And return step 2; When the number of times of images acquired from video data surpasses preset number of times, when still not having the alternative tracking target of coupling, then withdraw from this testing process.
5. a method that adopts the described object detection method of claim 1 to realize target following is characterized in that said method comprises the steps:
The target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of said tracking target area piece for preset number;
Judge the target that whether has said tracking in the image of subsequent frame according to the characteristic information of said target area piece, the output tracking result.
6. method as claimed in claim 5 is characterized in that, said certain orientation is vertical direction or horizontal direction.
7. method as claimed in claim 5 is characterized in that, the characteristic information of said target area piece is one or more in the relative position information between the color of target area piece, shape, texture and the target area piece.
8. an object detection system is characterized in that, said system comprises video data input block, ratio acquiring unit, ratio in judgement unit, threshold adjustment unit and object detection unit, wherein:
The video data input block is used to receive the video data that client is imported, and gathers any two field picture in the said video data, and pixel threshold is set;
The ratio acquiring unit; Be used for pixel threshold according to the setting of video data input block; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image; Be specially the absolute value of the background image respective point pixel intensity value difference that obtains said image and preset; Said preset background image is not comprising of obtaining in advance of the image during target to be detected; Pixel in the corresponding said image is confirmed as the pixel of target during greater than said pixel threshold with absolute value; Obtain the ratio of pixel number total in pixel number and the said image of target in the said image, perhaps in the output result of said ratio in judgement unit for not the time, obtain the ratio of pixel number total in pixel number and the said image of target in the said image according to the pixel threshold of said threshold adjustment unit output;
The ratio in judgement unit, whether the ratio that is used to judge the output of ratio acquiring unit is in preset scope;
Threshold adjustment unit is used in the output result of ratio in judgement unit for not the time adjustment pixel threshold; And
Object detection unit is used in the output result of ratio in judgement unit when being, obtains the pixel attribute information of target in the image according to pixel threshold, obtains the target of tracking according to said pixel attribute information.
9. a Target Tracking System that comprises the described object detection system of claim 8 is characterized in that, said system also comprises:
Target area piece acquiring unit is used for the target of the tracking of obtaining is divided into the identical object block of size according to certain orientation, according to the color similarity degree magnitude relationship between the object block, with the target identification of the said tracking target area piece for preset number; And
The tracking results output unit is used for judging according to the characteristic information of said target area piece whether the image of subsequent frame exists the target of said tracking, output tracking result.
10. a video monitoring equipment is characterized in that, said equipment comprises described object detection system of claim 8 or the described Target Tracking System of claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100694470A CN102136147B (en) | 2011-03-22 | 2011-03-22 | Target detecting and tracking method, system and video monitoring device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100694470A CN102136147B (en) | 2011-03-22 | 2011-03-22 | Target detecting and tracking method, system and video monitoring device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102136147A CN102136147A (en) | 2011-07-27 |
| CN102136147B true CN102136147B (en) | 2012-08-22 |
Family
ID=44295921
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011100694470A Expired - Fee Related CN102136147B (en) | 2011-03-22 | 2011-03-22 | Target detecting and tracking method, system and video monitoring device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102136147B (en) |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104298955A (en) * | 2013-07-15 | 2015-01-21 | 深圳市振邦实业有限公司 | Human head detection method and device |
| CN104424633B (en) * | 2013-08-23 | 2018-01-12 | 浙江大华技术股份有限公司 | A kind of video contrast's method for detecting abnormality and device |
| CN104202559A (en) * | 2014-08-11 | 2014-12-10 | 广州中大数字家庭工程技术研究中心有限公司 | Intelligent monitoring system and intelligent monitoring method based on rotation invariant feature |
| CN107491714B (en) * | 2016-06-13 | 2022-04-05 | 中科晶锐(苏州)科技有限公司 | Intelligent robot and target object identification method and device thereof |
| CN107564071A (en) * | 2016-07-01 | 2018-01-09 | 苏州宝时得电动工具有限公司 | A kind of image recognition meadow method and device |
| JP6776719B2 (en) * | 2016-08-17 | 2020-10-28 | 富士通株式会社 | Mobile group detection program, mobile group detection device, and mobile group detection method |
| CN106791715A (en) * | 2017-02-24 | 2017-05-31 | 深圳英飞拓科技股份有限公司 | Classification joint control intelligent control method and system |
| CN107944337A (en) * | 2017-10-13 | 2018-04-20 | 西安天和防务技术股份有限公司 | A kind of low target intelligent-tracking method and system, storage medium and electric terminal |
| CN109882761A (en) * | 2017-12-06 | 2019-06-14 | 爱玛丽欧股份有限公司 | The headlamp and its control method with tracking are automatically controlled using image identification |
| CN109284673B (en) * | 2018-08-07 | 2022-02-22 | 北京市商汤科技开发有限公司 | Object tracking method and device, electronic equipment and storage medium |
| CN111046788A (en) * | 2019-12-10 | 2020-04-21 | 北京文安智能技术股份有限公司 | Method, device and system for detecting staying personnel |
| CN112333417A (en) * | 2020-04-30 | 2021-02-05 | 深圳Tcl新技术有限公司 | Doorbell disturbance-free mode starting method and device and computer-readable storage medium |
| CN112051216B (en) * | 2020-05-20 | 2023-04-28 | 广东轻工职业技术学院 | Fruit dynamic tracking method for realizing fruit grading sorting |
| CN111845557B (en) * | 2020-07-23 | 2022-04-29 | 深圳市健创电子有限公司 | Safety early warning method and system for vehicle driving and related device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101236656A (en) * | 2008-02-29 | 2008-08-06 | 上海华平信息技术股份有限公司 | Movement target detection method based on block-dividing image |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3134845B2 (en) * | 1998-07-03 | 2001-02-13 | 日本電気株式会社 | Apparatus and method for extracting object in moving image |
| US7940961B2 (en) * | 2007-12-20 | 2011-05-10 | The United States Of America As Represented By The Secretary Of The Navy | Method for enhancing ground-based detection of a moving object |
-
2011
- 2011-03-22 CN CN2011100694470A patent/CN102136147B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101236656A (en) * | 2008-02-29 | 2008-08-06 | 上海华平信息技术股份有限公司 | Movement target detection method based on block-dividing image |
Non-Patent Citations (4)
| Title |
|---|
| Jong Bae Kim, et al..A Real-Time Region-Based Motion Segmentation Using Adaptive Thresholding and K-Means Clustering.《AI 2001: ADVANCES IN ARTIFICIAL INTELLIGENCE Lecture Notes in Computer Science》.2001,第2256卷213-224. * |
| JP特開2000-20722A 2000.01.21 |
| 王明吉,等..抗干扰的运动目标检测算法.《计算机工程》.2010,第36卷(第3期),232-233. * |
| 王明吉，等。.抗干扰的运动目标检测算法.《计算机工程》.2010,第36卷(第3期),232-233. |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102136147A (en) | 2011-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102136147B (en) | Target detecting and tracking method, system and video monitoring device | |
| Park et al. | Construction worker detection in video frames for initializing vision trackers | |
| CN107016367B (en) | Tracking control method and tracking control system | |
| US9652863B2 (en) | Multi-mode video event indexing | |
| US9792664B2 (en) | System and method for mapping object coordinates from a video to real world coordinates using perspective transformation | |
| CN111242025B (en) | Real-time action monitoring method based on YOLO | |
| CN102447835A (en) | Non-blind-area multi-target cooperative tracking method and system | |
| CN102982341A (en) | Self-intended crowd density estimation method for camera capable of straddling | |
| CN101344965A (en) | Tracking system based on binocular camera shooting | |
| CN103581614A (en) | Method and system for tracking targets in video based on PTZ | |
| TWI694352B (en) | Interactive behavior detection method, device, system and equipment | |
| Park et al. | Detection of construction workers in video frames for automatic initialization of vision trackers | |
| CN112184773A (en) | Helmet wearing detection method and system based on deep learning | |
| CN103577832B (en) | A kind of based on the contextual people flow rate statistical method of space-time | |
| CN111627049B (en) | Method and device for determining high-altitude parabolic object, storage medium and processor | |
| EP3051495B1 (en) | Systems and methods for mapping object coordinates from a video frame view to real world coordinates | |
| CN103607558A (en) | Video monitoring system, target matching method and apparatus thereof | |
| CN108010058A (en) | A kind of method and system that vision tracking is carried out to destination object in video flowing | |
| CN102663350A (en) | Road tunnel fire detection method based on video | |
| CN114220044A (en) | River course floater detection method based on AI algorithm | |
| CN106558069A (en) | A kind of method for tracking target and system based under video monitoring | |
| Yun et al. | Maximum-likelihood object tracking from multi-view video by combining homography and epipolar constraints | |
| JP6831396B2 (en) | Video monitoring device | |
| Park | Automated 3 D vision-based tracking of construction entities | |
| CN111985331A (en) | Detection method and device for preventing secret of business from being stolen |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120822 |