CN103810696B - Method for detecting image of target object and device thereof - Google Patents
Method for detecting image of target object and device thereof Download PDFInfo
- Publication number
- CN103810696B CN103810696B CN201210463919.5A CN201210463919A CN103810696B CN 103810696 B CN103810696 B CN 103810696B CN 201210463919 A CN201210463919 A CN 201210463919A CN 103810696 B CN103810696 B CN 103810696B
- Authority
- CN
- China
- Prior art keywords
- image
- region
- destination object
- images match
- setting regions
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for detecting the image of a target object. The method comprises the following steps: obtaining the coordinate position of an image region in the last frame of the image when the image region which is matched with the image of a target object exists in the preset region of the last frame of the image of the current frame of the image; detecting whether the image region which is matched with the image of the target object exists in a first set region of the obtained coordinate position contained in the current frame of the image, wherein the first set region is located in the preset region; determining that the image region which is matched with the image of the target object exists in the preset region of the current frame of the image when the image region which is matched with the image of the target object exists in the first set region. Through the adoption of the method and the device provided by the invention, the problem of lower efficiency in the process of detecting the image of the target object is solved.
Description
Technical field
The present invention relates to field of video monitoring, more particularly to a kind of destination object image detecting method and device.
Background technology
Video monitoring is the important component part of safety and protection system, is early widely used to many occasions of every profession and trade.
Video monitoring presents more and more intelligentized development trend after it experienced digitized and networking, at present.
In many video monitoring scenes, it is required for destination object, such as people, animal, vehicle etc. are with the presence or absence of in default
Region carries out detecting, tracks and behavior analysiss.In prison, specific operator on duty is needed to be stood guard or gone on patrol, if on duty
Personnel leave the post without authorization, it is impossible to be normally carried out the process of event, then may cause uncontrollable accident, accordingly, it would be desirable to value
Whether personnel are on duty is detected for class.
The scheme for generally adopting in the prior art is the panorama sketch to the predeterminable area in every frame video image for obtaining
Detected as interior, it is determined whether there is destination object, although the program can be realized to destination object with the presence or absence of in default
Detected in region, but detected in the panoramic picture of the predeterminable area in every two field picture, due to detection zone it is larger,
So less efficient.
The content of the invention
The embodiment of the present invention provides a kind of destination object image detecting method and device, to solve presence in prior art
The problem less efficient when detecting to destination object image.
The embodiment of the present invention provides a kind of destination object image detecting method, including:
When the image-region of presence and destination object images match in the predeterminable area of the previous frame image of current frame image
When, obtain coordinate position of the described image region in the previous frame image;
Detect whether exist in the first setting regions comprising the coordinate position for obtaining in the current frame image
With the image-region of the destination object images match;First setting regions is located in the predeterminable area;
When there is the image-region with the destination object images match in first setting regions, it is determined that described work as
There is the image-region with the destination object images match in the predeterminable area of prior image frame.
The embodiment of the present invention provides a kind of destination object image detection device, including:
Acquiring unit, for when presence in the predeterminable area of the previous frame image of current frame image and destination object image
During the image-region matched somebody with somebody, coordinate position of the described image region in the previous frame image is obtained;
Detector unit, in the first setting regions comprising the coordinate position for obtaining in the current frame image
Detect whether the image-region for having with the destination object images match;First setting regions is located at the predeterminable area
It is interior;
Determining unit, for working as the image-region existed in first setting regions with the destination object images match
When, determine the image-region existed in the predeterminable area of the current frame image with the destination object images match.
Beneficial effect of the present invention includes:
Method provided in an embodiment of the present invention, if existed and destination object image in the previous frame image of current frame image
The image-region of matching, obtains its position coordinates, when the detection of destination object image is carried out in current frame image, should including
Detected in first setting regions of position coordinateses, as, in adjacent two field pictures, the position of destination object image may
Constant or change is less, that is, the probability occurred in the first setting regions is larger, and as the first setting regions is less than pre-
If region such that it is able to improve detection efficiency.
Description of the drawings
Accompanying drawing is used for providing a further understanding of the present invention, and constitutes a part for description, with present invention enforcement
Example is used for together explaining the present invention, is not construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the flow chart of destination object image detecting method provided in an embodiment of the present invention;
Fig. 2 is the detail flowchart of the destination object image detecting method that the embodiment of the present invention 1 is provided;
Fig. 3 is the detail flowchart of the destination object image detecting method that the embodiment of the present invention 2 is provided;
Fig. 4 is the detail flowchart of the destination object image detecting method that the embodiment of the present invention 3 is provided;
Fig. 5 is the structural representation of the destination object image detection device that the embodiment of the present invention 4 is provided.
Specific embodiment
In order to provide the implementation for improving destination object image detection efficiency, a kind of target is embodiments provided
The preferred embodiments of the present invention are illustrated by object images detection method and device below in conjunction with Figure of description, should be managed
Solution, preferred embodiment described herein are merely to illustrate and explain the present invention, be not intended to limit the present invention.And do not rushing
In the case of prominent, the feature in embodiment and embodiment in the application can be mutually combined.
The embodiment of the present invention provides a kind of destination object image detecting method, as shown in figure 1, including:
Step 101, when existing with destination object images match in the predeterminable area of the previous frame image of current frame image
During image-region, coordinate position of the image-region in previous frame image is obtained.
Detect whether to deposit in step 102, the first setting regions comprising the coordinate position for obtaining in current frame image
In the image-region with destination object images match;First setting regions is located in predeterminable area.
Step 103, when existing with the image-region of destination object images match in the first setting regions, determine present frame
There is the image-region with destination object images match in the predeterminable area of image.
Further, when there is no the image-region with destination object images match in the first setting regions, can be true
There is no the image-region with destination object images match in the predeterminable area of settled prior image frame.
However, when there is no the image-region with destination object images match in the first setting regions, directly determining and working as
There is no the scheme with the image-region of destination object images match in the predeterminable area of prior image frame, may affect to target
The accuracy of object images detection.
Preferably, when there is no the image-region with destination object images match in the first setting regions, can also sentence
Whether there is foreground image in the predeterminable area of disconnected current frame image;Before it is determined that not existing in the predeterminable area of current frame image
During scape image, in the predeterminable area for determining current frame image, there is no the image-region with destination object images match;When it is determined that
When there is foreground image in the predeterminable area of current frame image, the comprising the foreground image second setting area in current frame image
The image-region for having with destination object images match is detected whether in domain;Second setting regions is located in predeterminable area.
When there is the image-region with destination object images match in the second setting regions, the pre- of current frame image is determined
If there is the image-region with destination object images match in region;When not existing in the second setting regions and destination object image
During the image-region of matching, in the predeterminable area for determining current frame image, there is no the image district with destination object images match
Domain.
Above-mentioned steps implementing precondition in the inventive method is in the predeterminable area of the previous frame image of current frame image
Exist and the image-region of destination object images match, when do not exist in the predeterminable area of the previous frame image of current frame image with
During the image-region of destination object images match, directly detect whether exist and target pair in the predeterminable area of current frame image
As the image-region of images match.
In target area(Including the first setting regions, the second setting regions and predeterminable area)Inside detect whether exist and mesh
The image-region of mark object images matching specifically can be in the following way:
Mode one:It is more than texture phase with the presence or absence of with the gray scale texture similarity of destination object image in detection target area
Like the image-region for spending threshold value;It is similar more than texture to the gray scale texture similarity of destination object image when existing in target area
During the image-region of degree threshold value, the image-region existed in target area with destination object images match is determined;Work as target area
Inside do not exist with the gray scale texture similarity of destination object image more than texture similarity threshold value image-region when, determine target
There is no the image-region with destination object images match in region;
Mode two:It is more than image phase with the presence or absence of with the segmentation image similarity of destination object image in detection target area
Like the image-region for spending threshold value;When in target area, presence is more than image similarity with the segmentation image similarity of destination object image
During the image-region of degree threshold value, the image-region existed in target area with destination object images match is determined;Work as target area
Inside do not exist with destination object image segmentation image similarity more than image similarity threshold value image-region when, determine target
There is no the image-region with destination object images match in region;
Mode three:It is more than texture phase with the presence or absence of with the gray scale texture similarity of destination object image in detection target area
Like degree threshold value, and with destination object image segmentation image similarity more than image similarity threshold value image-region;Work as mesh
Exist in mark region with the gray scale texture similarity of destination object image more than texture similarity threshold value, and with destination object figure
When the segmentation image similarity of picture is more than the image-region of image similarity threshold value, determines in target area and exist and destination object
The image-region of images match;It is more than texture phase with the gray scale texture similarity of destination object image when not existing in target area
Like degree threshold value, and with destination object image segmentation image similarity more than image similarity threshold value image-region when, really
Set the goal and there is no image-region with destination object images match in region.
Based on the gray scale texture similarity of image, and/or segmentation image similarity detects whether exist in target area
Method with the image-region of destination object images match is only an example, and other detection methods also can be used as the tool of this step
Body implementation.
Further, the image-region for meeting multiple matching conditions can be defined as with destination object images match really
Determine image-region, only met one of the plurality of matching condition image-region be defined as it is potential with destination object images match
Image-region.
When the determination image of presence and destination object images match in the predeterminable area of the previous frame image of current frame image
During region, obtain this and determine coordinate position of the image-region in previous frame image;When the previous frame image of current frame image
There is no the determination image-region with destination object images match in predeterminable area, and the previous frame image of current frame image is pre-
If there is the latent image region with destination object images match in region, the latent image region is obtained in previous frame image
In coordinate position.
Detect whether exist and target in the first setting regions comprising the coordinate position for obtaining in current frame image
The determination image-region of object images matching, and with the presence or absence of the latent image region with destination object images match;When
When there is the determination image-region with destination object images match in one setting regions, determine in the predeterminable area of current frame image
There is the determination image-region with destination object images match;When presence in the first setting regions and destination object images match
During latent image region, the latent image area existed in the predeterminable area of current frame image with destination object images match is determined
Domain.
Below in conjunction with the accompanying drawings, the method and device that the present invention is provided is described in detail with specific embodiment.
Embodiment 1:
Fig. 2 show the detail flowchart of the destination object image detection of the offer of the embodiment of the present invention 1, specifically includes as follows
Process step:
Step 201, judge current frame image previous frame image predeterminable area in the presence or absence of with destination object image
The image-region of matching.
When it is determined that current frame image previous frame image predeterminable area in there is no figure with destination object images match
During as region, into step 202;When it is determined that current frame image previous frame image predeterminable area in exist with destination object figure
During the image-region of picture matching, into step 203.
Step 202, directly detect whether exist and destination object images match in the predeterminable area of current frame image
Image-region.
The first setting regions in predeterminable area and subsequent step, the second setting in the present embodiment, in this step 202
Aforesaid way three is adopted when detecting whether to exist in the target areas such as region the image-region with destination object images match specifically:
It is more than texture similarity threshold with the presence or absence of with the gray scale texture similarity of destination object image in detection target area
Value, and with destination object image segmentation image similarity more than image similarity threshold value image-region;Work as target area
Interior presence and the gray scale texture similarity of destination object image are more than texture similarity threshold value, and with destination object image point
Cut image similarity more than image similarity threshold value image-region when, determine the target area in exist and destination object image
The image-region matched somebody with somebody;It is more than texture similarity threshold with the gray scale texture similarity of destination object image when not existing in target area
Value, and with destination object image segmentation image similarity more than image similarity threshold value image-region when, determine target
There is no the image-region with destination object images match in region.
Specifically, LBP can be based on(LocalBinaryPattern, local binary patterns)Numerical value, carries out gray scale texture phase
Like the calculating of degree;Split based on grey level histogram, carry out splitting the calculating of image similarity.
Step 207 or step 208 are directly entered according to testing result.
Step 203, obtain coordinate position of the image-region in previous frame image.
Detect whether to deposit in step 204, the first setting regions comprising the coordinate position for obtaining in current frame image
In the image-region with destination object images match;First setting regions is located in predeterminable area.
When it is determined that not existing and target pair in the first setting regions of the coordinate position for including acquisition in current frame image
As images match image-region when, into step 205;When it is determined that comprising the coordinate position for obtaining in current frame image the
When there is the image-region with destination object images match in one setting regions, into step 207.
Step 205, judge in the predeterminable area of current frame image with the presence or absence of foreground image.
When it is determined that there is foreground image in the predeterminable area of current frame image, into step 206;When it is determined that present frame figure
When there is no foreground image in the predeterminable area of picture, into step 208.
In the present embodiment, background image can be generated using any background modeling method, and background image also constantly will enter
Row study updates.According to background image and the predeterminable area of current frame image, whether judge in the predeterminable area of current frame image
There is foreground image.
In step 206, the second setting regions for the foreground image being included in current frame image detect whether exist and mesh
The image-region of mark object images matching;Second setting regions is located in predeterminable area.
When it is determined that existing and destination object image in the second setting regions comprising the foreground image in current frame image
During the image-region matched somebody with somebody, into step 207;When it is determined that in current frame image in the second setting regions comprising the foreground image not
When there is the image-region with destination object images match, into step 208.
Preferably, can be detected in the region beyond the first setting regions in the second setting regions.
Step 207, the image-region for determining presence and destination object images match in the predeterminable area of current frame image.
Terminate the testing process in current frame image, and record coordinate bit of the image-region in current frame image
Put, use during for the destination object image being detected in the predeterminable area of next two field picture of current frame image.
Step 208, the interior image-region not existed with destination object images match of predeterminable area for determining current frame image.
By when in above-described embodiment 1, method is applied to the detection on duty of operator on duty, the image of operator on duty is as above-mentioned
Destination object image in embodiment 1, the image of the operator on duty can be stored in advance, it is also possible on duty in operator on duty
It is configured in real time afterwards;Corresponding region is presetting in above-described embodiment 1 to the movable scope of operator on duty in the picture
Region, the movable scope of the operator on duty for standing guard are less, and corresponding predeterminable area is also less, the movable model of operator on duty of patrol
Enclose larger, corresponding predeterminable area is also larger.
It should be noted that background image is that constantly study updates, the actionless object of long-time will be learned to
In background image.When 1 method of embodiment is applied to the detection on duty of operator on duty, no matter stood guard due to operator on duty or
Patrol, all there may be that operator on duty is motionless or the situation of microinching, therefore should reduce the pace of learning of background image, to carry
The high accuracy to operator on duty's image detection.
Also, in actual applications, monitoring scene is complicated, it is impossible to judge on duty only according to the testing result of a two field picture
Whether personnel leave the post, and can preserve the testing result of each two field picture, if continuously setting in the predeterminable area of number of frames picture
When there is no operator on duty's image, just determine that operator on duty leaves the post, and sends report of leaving the post, start and report to the police.
The method provided using the embodiment of the present invention 1, according to the image-region of destination object images match current
Position coordinateses in the previous frame image of two field picture, are carried out in the less region comprising the position coordinateses in current frame image
The detection of destination object image, it is possible to increase detection efficiency;And when in the region comprising the position coordinateses of current frame image
When there is no the image-region with destination object images match, further detected in the foreground image of predeterminable area, energy
Enough accuracy for ensureing detection.
Preferably, will can also exist in target area that texture is more than with the gray scale texture similarity of destination object image
Similarity threshold, but the image-region with the segmentation image similarity no more than image similarity threshold value of destination object image, or
It is existing in person target area to be not more than texture similarity threshold value, but and target with destination object image gray scale texture similarity
The segmentation image similarity of object images is defined as latent image region more than the image-region of image similarity threshold value;By target
What is existed in region is more than texture similarity threshold value with the gray scale texture similarity of destination object image, and with destination object figure
The segmentation image similarity of picture is defined as determining image-region more than the image-region of image similarity threshold value.Now, accordingly
The handling process detailed in Example 2 of destination object image detection.
Embodiment 2:
Fig. 3 show the detail flowchart of the destination object image detection of the offer of the embodiment of the present invention 2, specifically includes as follows
Process step:
Step 301, judge current frame image previous frame image predeterminable area in the presence or absence of with destination object image
The determination image-region of matching.
When it is determined that current frame image previous frame image predeterminable area in do not exist with destination object images match really
When determining image-region, into step 302;When it is determined that current frame image previous frame image predeterminable area in exist with target pair
As images match determination image-region when, into step 304.
Step 302, judge current frame image previous frame image predeterminable area in whether do not exist and destination object figure
As the latent image region of matching.
When it is determined that current frame image previous frame image predeterminable area in do not exist it is latent with destination object images match
In image-region, into step 303;When it is determined that current frame image previous frame image predeterminable area in exist with target pair
As images match latent image region when, into step 307.
Step 303, directly detect whether exist and destination object images match in the predeterminable area of current frame image
Determine image-region, and with the presence or absence of the latent image region with destination object images match.
Step 309, step 310 or step 311 are directly entered according to testing result.
Step 304, obtain this and determine coordinate position of the image-region in previous frame image.
Step 305, in current frame image comprising determine coordinate position of the image-region in previous frame image first
The determination image-region for having with destination object images match is detected whether in setting regions, and be whether there is and destination object
The latent image region of images match, first setting regions are located in predeterminable area.
When it is determined that there is no the determination image-region with destination object images match in first setting regions, into step
Rapid 306;When it is determined that there is the determination image-region with destination object images match in first setting regions, into step
309;When it is determined that there is the latent image region with destination object images match in first setting regions, into step 311.
Step 306, judge current frame image previous frame image predeterminable area in the presence or absence of with destination object image
The latent image region of matching.
When it is determined that current frame image previous frame image predeterminable area in exist it is potential with destination object images match
During image-region, into step 307;When it is determined that current frame image previous frame image predeterminable area in do not exist and target pair
As images match latent image region when, into step 310.
Step 307, obtain the coordinate position of the latent image region in previous frame image.
Step 308, in current frame image comprising coordinate position of the latent image region in previous frame image first
The determination image-region for having with destination object images match is detected whether in setting regions, and be whether there is and destination object
The latent image region of images match, first setting regions are located in predeterminable area.
When it is determined that there is the determination image-region with destination object images match in first setting regions, into step
309;When it is determined that there is no the determination image-region with destination object images match in first setting regions, into step
310;When it is determined that there is the latent image region with destination object images match in first setting regions, into step 311.
Step 309, the determination image district for determining presence and destination object images match in the predeterminable area of current frame image
Domain.
Step 310, the interior determination image not existed with destination object images match of predeterminable area for determining current frame image
Region.
Step 311, the latent image area for determining presence and destination object images match in the predeterminable area of current frame image
Domain.
Preferably, continuously ought set in the predeterminable area of quantity two field picture not existing with destination object images match really
When determining image-region, start and report to the police.
The method provided using the embodiment of the present invention 2, for present frame picture, carries out target pair in less region
As the detection of image, it is possible to increase detection efficiency, and comprising coordinate position of the latent image region in previous frame image
The first setting regions in be also carried out detection, relative to only comprising determine coordinate position of the image-region in previous frame image
The first setting regions in the scheme that detected, it is possible to increase the accuracy of detection.
On the basis of efficiency is improved, in order to further improve the accuracy of destination object image detection, can also adopt
Method in following embodiments 3.
Embodiment 3:
Fig. 4 show the detail flowchart of the destination object image detection of the offer of the embodiment of the present invention 3, specifically includes as follows
Process step:
Step 401, judge current frame image previous frame image predeterminable area in the presence or absence of with destination object image
The determination image-region of matching.
When it is determined that current frame image previous frame image predeterminable area in do not exist with destination object images match really
When determining image-region, into step 402;When it is determined that current frame image previous frame image predeterminable area in exist with target pair
As images match determination image-region when, into step 404.
Step 402, judge current frame image previous frame image predeterminable area in whether do not exist and destination object figure
As the latent image region of matching.
When it is determined that current frame image previous frame image predeterminable area in do not exist it is latent with destination object images match
In image-region, into step 403;When it is determined that current frame image previous frame image predeterminable area in exist with target pair
As images match latent image region when, into step 407.
Step 403, directly detect whether exist and destination object images match in the predeterminable area of current frame image
Determine image-region, and with the presence or absence of the latent image region with destination object images match.
Step 411, step 412 or step 413 are directly entered according to testing result.
Step 404, obtain this and determine coordinate position of the image-region in previous frame image.
Step 405, in current frame image comprising determine coordinate position of the image-region in previous frame image first
The determination image-region for having with destination object images match is detected whether in setting regions, and be whether there is and destination object
The latent image region of images match, first setting regions are located in predeterminable area.
When it is determined that there is no the determination image-region with destination object images match in first setting regions, into step
Rapid 406;When it is determined that there is the determination image-region with destination object images match in first setting regions, into step
411;When it is determined that there is the latent image region with destination object images match in first setting regions, into step 413.
Step 406, judge current frame image previous frame image predeterminable area in the presence or absence of with destination object image
The latent image region of matching.
When it is determined that current frame image previous frame image predeterminable area in exist it is potential with destination object images match
During image-region, into step 407;When it is determined that current frame image previous frame image predeterminable area in do not exist and target pair
As images match latent image region when, into step 412.
Step 407, obtain the coordinate position of the latent image region in previous frame image.
Step 408, in current frame image comprising coordinate position of the latent image region in previous frame image first
The determination image-region for having with destination object images match is detected whether in setting regions, and be whether there is and destination object
The latent image region of images match, first setting regions are located in predeterminable area.
When it is determined that there is no the determination image-region with destination object images match in first setting regions, into step
Rapid 409;When it is determined that there is the determination image-region with destination object images match in first setting regions, into step
411;When it is determined that there is the latent image region with destination object images match in first setting regions, into step 413.
Step 409, judge in the predeterminable area of current frame image with the presence or absence of foreground image.
When it is determined that there is foreground image in the predeterminable area of current frame image, into step 410;When it is determined that present frame figure
When there is no foreground image in the predeterminable area of picture, into step 412.
In step 410, the second setting regions for the foreground image being included in current frame image detect whether exist and mesh
The determination image-region of mark object images matching, and with the presence or absence of the latent image region with destination object images match;Should
Second setting regions is located in predeterminable area.
When it is determined that there is the determination image-region with destination object images match in second setting regions, into step
411;When it is determined that there is no the image-region with destination object images match in second setting regions, into step 412;When
When determining the latent image region of presence and destination object images match in second setting regions, into step 413.
Preferably, can be detected in the region beyond the first setting regions in the second setting regions.
Step 411, the determination image district for determining presence and destination object images match in the predeterminable area of current frame image
Domain.
Step 412, the interior determination image not existed with destination object images match of predeterminable area for determining current frame image
Region.
Step 413, the latent image area for determining presence and destination object images match in the predeterminable area of current frame image
Domain.
Preferably, continuously ought set in the predeterminable area of quantity two field picture not existing with destination object images match really
When determining image-region, start and report to the police.
The method provided using the embodiment of the present invention 3, relative to the method provided by embodiment 2, further default
Detected in the foreground image in region, the accuracy of detection can be further improved on the basis of detection efficiency is improved.
Embodiment 4:
Based on same inventive concept, according to the destination object image detecting method method that the above embodiment of the present invention is provided,
Correspondingly, another embodiment of the present invention additionally provides destination object image detection device, apparatus structure schematic diagram as shown in figure 5,
Specifically include:
Acquiring unit 501, for when presence in the predeterminable area of the previous frame image of current frame image and destination object figure
During the image-region of picture matching, coordinate position of the image-region in previous frame image is obtained;
Detector unit 502, for examining in the first setting regions comprising the coordinate position for obtaining in current frame image
Survey with the presence or absence of the image-region with destination object images match;First setting regions is located in predeterminable area;
Determining unit 503, for when existing with the image-region of destination object images match in the first setting regions, really
There is the image-region with destination object images match in the predeterminable area of settled prior image frame.
Further, determining unit 503, are additionally operable to not exist and destination object images match in the first setting regions
Image-region, when there is no foreground image in the predeterminable area of current frame image, determines in the predeterminable area of current frame image not
There is the image-region with destination object images match.
Further, detector unit 502, are additionally operable to not exist and destination object images match in the first setting regions
Image-region, when there is foreground image in the predeterminable area of current frame image, includes the foreground image in current frame image
The image-region for having with destination object images match is detected whether in second setting regions;Second setting regions is located at default
In region;When there is the image-region with destination object images match in the second setting regions, the pre- of current frame image is determined
If there is the image-region with destination object images match in region.
Further, determining unit 503, are additionally operable to not exist and destination object images match in the second setting regions
During image-region, in the predeterminable area for determining current frame image, there is no the image-region with destination object images match.
Further, detector unit 502, specifically for detect the first setting regions in the presence or absence of with destination object image
Gray scale texture similarity more than texture similarity threshold value image-region;When presence in the first setting regions and destination object figure
When the gray scale texture similarity of picture is more than the image-region of texture similarity threshold value, determines in the first setting regions and exist and target
The image-region of object images matching;It is big with the gray scale texture similarity of destination object image when not existing in the first setting regions
When the image-region of texture similarity threshold value, the image not existed in the first setting regions with destination object images match is determined
Region;Or
Detect in the first setting regions and be more than image similarity with the presence or absence of with the segmentation image similarity of destination object image
The image-region of degree threshold value;When in the first setting regions, presence is more than image phase with the segmentation image similarity of destination object image
During like spending the image-region of threshold value, the image-region existed in the first setting regions with destination object images match is determined;When
There is no the image-region more than image similarity threshold value with the segmentation image similarity of destination object image in one setting regions
When, determine the image-region not existed in the first setting regions with destination object images match;Or
Detect in the first setting regions with the presence or absence of similar more than texture to the gray scale texture similarity of destination object image
Degree threshold value, and with destination object image segmentation image similarity more than image similarity threshold value image-region;When first
Exist in setting regions and the gray scale texture similarity of destination object image is more than texture similarity threshold value, and and destination object
When the segmentation image similarity of image is more than the image-region of image similarity threshold value, determines in the first setting regions and exist and mesh
The image-region of mark object images matching;When there is no the gray scale texture similarity with destination object image in the first setting regions
More than texture similarity threshold value, and with destination object image segmentation image similarity more than image similarity threshold value image
During region, the image-region not existed in the first setting regions with destination object images match is determined.
Further, acquiring unit 501, specifically for when presence in the predeterminable area of the previous frame image of current frame image
During with the determination image-region of destination object images match, obtain this and determine coordinate bit of the image-region in previous frame image
Put;When there is no the determination image district with destination object images match in the predeterminable area of the previous frame image of current frame image
There is the latent image region with destination object images match in domain, and the predeterminable area of the previous frame image of current frame image
When, obtain the coordinate position of the latent image region in previous frame image;Wherein, the determination figure with destination object images match
As region is the image-region that meets multiple matching conditions, should only to meet with the latent image region of destination object images match
The image-region of one of multiple matching conditions;
Detector unit 502, specifically for the first setting regions comprising the coordinate position for obtaining in current frame image
Inside detect whether exist and the determination image-region of destination object images match, and with the presence or absence of with destination object images match
Latent image region;
Determining unit 503, specifically for when the determination image of presence and destination object images match in the first setting regions
During region, the determination image-region existed in the predeterminable area of current frame image with destination object images match is determined;When first
When there is the latent image region with destination object images match in setting regions, the predeterminable area internal memory of current frame image is determined
In the latent image region with destination object images match.
In sum, scheme provided in an embodiment of the present invention, when in the predeterminable area of the previous frame image of current frame image
When there is the image-region with destination object images match, coordinate position of the image-region in previous frame image is obtained;
Detect whether exist and destination object image in the first setting regions comprising the coordinate position for obtaining in current frame image
The image-region matched somebody with somebody;First setting regions is located in predeterminable area;When presence in the first setting regions and destination object image
During the image-region of matching, the image-region existed in the predeterminable area of current frame image with destination object images match is determined.
The scheme provided using the present invention, is solved the problems, such as less efficient when detecting to destination object image.
The destination object image detection device provided by embodiments herein can be realized by computer program.This area
Technical staff is it should be appreciated that above-mentioned Module Division mode is only the one kind in numerous Module Division modes, if divided
For other modules or non-division module, as long as destination object image detection device has above-mentioned functions, all should be the application's
Within protection domain.
The application is with reference to method, the equipment according to the embodiment of the present application(System), and computer program flow process
Figure and/or block diagram are describing.It should be understood that can be by computer program instructions flowchart and/or each stream in block diagram
The combination of journey and/or square frame and flow chart and/or flow process and/or square frame in block diagram.These computer programs can be provided
The processor of general purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced for reality by the instruction of computer or the computing device of other programmable data processing devices
The device of the function of specifying in present one flow process of flow chart or one square frame of multiple flow processs and/or block diagram or multiple square frames.
These computer program instructions may be alternatively stored in and can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory is produced to be included referring to
Make the manufacture of device, the command device realize in one flow process of flow chart or one square frame of multiple flow processs and/or block diagram or
The function of specifying in multiple square frames.
These computer program instructions can be also loaded in computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented process, so as in computer or
The instruction performed on other programmable devices is provided for realizing in one flow process of flow chart or multiple flow processs and/or block diagram one
The step of function of specifying in individual square frame or multiple square frames.
Obviously, those skilled in the art can carry out the essence of various changes and modification without deviating from the present invention to the present invention
God and scope.So, if these modifications of the present invention and modification belong to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to comprising these changes and modification.
Claims (12)
1. a kind of destination object image detecting method, it is characterised in that include:
When there is the image-region with destination object images match in the predeterminable area of the previous frame image of current frame image, obtain
Take coordinate position of the described image region in the previous frame image;
Detect whether exist and institute in the first setting regions comprising the coordinate position for obtaining in the current frame image
State the image-region of destination object images match;First setting regions is located in the predeterminable area;
When there is the image-region with the destination object images match in first setting regions, the present frame is determined
There is the image-region with the destination object images match in the predeterminable area of image.
2. the method for claim 1, it is characterised in that also include:
When there is no the image-region with the destination object images match, the current frame image in first setting regions
Predeterminable area in when there is no foreground image, do not exist and the target pair in the predeterminable area for determining the current frame image
As the image-region of images match.
3. method as claimed in claim 2, it is characterised in that also include:
When there is no the image-region with the destination object images match, the current frame image in first setting regions
Predeterminable area in when there is foreground image, in the current frame image in the second setting regions comprising the foreground image
Detect whether the image-region for having with the destination object images match;Second setting regions is located at the predeterminable area
It is interior;
When there is the image-region with the destination object images match in second setting regions, the present frame is determined
There is the image-region with the destination object images match in the predeterminable area of image.
4. method as claimed in claim 3, it is characterised in that also include:
When there is no the image-region with the destination object images match in second setting regions, determine described current
There is no the image-region with the destination object images match in the predeterminable area of two field picture.
5. the method for claim 1, it is characterised in that concrete to examine in first setting regions in the following way
Survey with the presence or absence of the image-region with the destination object images match:
Detect in first setting regions and be more than texture with the presence or absence of with the gray scale texture similarity of the destination object image
The image-region of similarity threshold;It is similar to the gray scale texture of the destination object image when existing in first setting regions
When degree is more than the image-region of texture similarity threshold value, determines in first setting regions and exist and the destination object image
The image-region of matching;It is big with the gray scale texture similarity of the destination object image when not existing in first setting regions
When the image-region of texture similarity threshold value, determine and do not exist and the destination object image in first setting regions
The image-region matched somebody with somebody;Or
Detect in first setting regions and be more than image with the presence or absence of with the segmentation image similarity of the destination object image
The image-region of similarity threshold;When there is the segmentation image similarity with the destination object image in first setting regions
When degree is more than the image-region of image similarity threshold value, determines in first setting regions and exist and the destination object image
The image-region of matching;It is big with the segmentation image similarity of the destination object image when not existing in first setting regions
When the image-region of image similarity threshold value, determine and do not exist and the destination object image in first setting regions
The image-region matched somebody with somebody;Or
Detect in first setting regions and be more than texture with the presence or absence of with the gray scale texture similarity of the destination object image
Similarity threshold, and with the destination object image segmentation image similarity more than image similarity threshold value image district
Domain;It is more than texture similarity threshold with the gray scale texture similarity of the destination object image when existing in first setting regions
Value, and with the destination object image segmentation image similarity more than image similarity threshold value image-region when, it is determined that
There is the image-region with the destination object images match in first setting regions;When in first setting regions not
Exist and be more than texture similarity threshold value with the gray scale texture similarity of the destination object image, and with the destination object figure
When the segmentation image similarity of picture is more than the image-region of image similarity threshold value, determines and do not exist in first setting regions
With the image-region of the destination object images match.
6. the method for claim 1, it is characterised in that when the predeterminable area internal memory of the previous frame image of current frame image
In the image-region with destination object images match, coordinate bit of the described image region in the previous frame image is obtained
Put, specifically include:
When the determination image-region of presence and destination object images match in the predeterminable area of the previous frame image of current frame image
When, obtain the determination coordinate position of the image-region in the previous frame image;When the previous frame image of current frame image
Predeterminable area in there is no determination image-region with destination object images match, and the previous frame image of current frame image
When there is the latent image region with destination object images match in predeterminable area, the latent image region is obtained on described
Coordinate position in one two field picture;
Wherein, described is the image-region for meeting multiple matching conditions with destination object images match determination image-region, institute
It is the image-region that only meets one of the plurality of matching condition to state with the latent image region of destination object images match;
Detect whether exist and institute in the first setting regions comprising the coordinate position for obtaining in the current frame image
The image-region of destination object images match is stated, is specifically included:
Detect whether exist and institute in the first setting regions comprising the coordinate position for obtaining in the current frame image
The determination image-region of destination object images match is stated, and with the presence or absence of the latent image with the destination object images match
Region;
When there is the image-region with the destination object images match in first setting regions, the present frame is determined
There is the image-region with the destination object images match in the predeterminable area of image, specifically include:
When there is the determination image-region with the destination object images match in first setting regions, it is determined that described work as
There is the determination image-region with the destination object images match in the predeterminable area of prior image frame;
When there is the latent image region with the destination object images match in first setting regions, it is determined that described work as
There is the latent image region with the destination object images match in the predeterminable area of prior image frame.
7. a kind of destination object image detection device, it is characterised in that include:
Acquiring unit, for when presence in the predeterminable area of the previous frame image of current frame image and destination object images match
During image-region, coordinate position of the described image region in the previous frame image is obtained;
Detector unit, for detecting in the first setting regions comprising the coordinate position for obtaining in the current frame image
With the presence or absence of the image-region with the destination object images match;First setting regions is located in the predeterminable area;
Determining unit, for when there is the image-region with the destination object images match in first setting regions,
Determine the image-region existed in the predeterminable area of the current frame image with the destination object images match.
8. device as claimed in claim 7, it is characterised in that the determining unit, is additionally operable to when first setting regions
Inside there is no the image-region with the destination object images match, in the predeterminable area of the current frame image, there is no prospect
During image, in the predeterminable area for determining the current frame image, there is no the image-region with the destination object images match.
9. device as claimed in claim 8, it is characterised in that the detector unit, is additionally operable to when first setting regions
Inside there is no the image-region with the destination object images match, in the predeterminable area of the current frame image, there is foreground picture
During picture, detect whether exist and the target in the second setting regions comprising the foreground image in the current frame image
The image-region of object images matching;Second setting regions is located in the predeterminable area;When second setting regions
When interior presence is with the image-region of the destination object images match, determine exist in the predeterminable area of the current frame image with
The image-region of the destination object images match.
10. device as claimed in claim 9, it is characterised in that the determining unit, is additionally operable to when second setting regions
When inside there is no the image-region with the destination object images match, do not deposit in the predeterminable area for determining the current frame image
In the image-region with the destination object images match.
11. devices as claimed in claim 7, it is characterised in that the detector unit, specifically for detecting first setting
With the presence or absence of the image-region for being more than texture similarity threshold value with the gray scale texture similarity of the destination object image in region;
It is more than texture similarity threshold value with the gray scale texture similarity of the destination object image when existing in first setting regions
Image-region when, determine exist in first setting regions and the destination object images match image-region;Work as institute
State and do not exist in the first setting regions with the gray scale texture similarity of the destination object image more than texture similarity threshold value
During image-region, the image-region not existed in first setting regions with the destination object images match is determined;Or
Detect in first setting regions and be more than image with the presence or absence of with the segmentation image similarity of the destination object image
The image-region of similarity threshold;When there is the segmentation image similarity with the destination object image in first setting regions
When degree is more than the image-region of image similarity threshold value, determines in first setting regions and exist and the destination object image
The image-region of matching;It is big with the segmentation image similarity of the destination object image when not existing in first setting regions
When the image-region of image similarity threshold value, determine and do not exist and the destination object image in first setting regions
The image-region matched somebody with somebody;Or
Detect in first setting regions and be more than texture with the presence or absence of with the gray scale texture similarity of the destination object image
Similarity threshold, and with the destination object image segmentation image similarity more than image similarity threshold value image district
Domain;It is more than texture similarity threshold with the gray scale texture similarity of the destination object image when existing in first setting regions
Value, and with the destination object image segmentation image similarity more than image similarity threshold value image-region when, it is determined that
There is the image-region with the destination object images match in first setting regions;When in first setting regions not
Exist and be more than texture similarity threshold value with the gray scale texture similarity of the destination object image, and with the destination object figure
When the segmentation image similarity of picture is more than the image-region of image similarity threshold value, determines and do not exist in first setting regions
With the image-region of the destination object images match.
12. devices as claimed in claim 7, it is characterised in that the acquiring unit, specifically for working as the upper of current frame image
When there is the determination image-region with destination object images match in the predeterminable area of one two field picture, the determination image district is obtained
Coordinate position of the domain in the previous frame image;When not existing and mesh in the predeterminable area of the previous frame image of current frame image
The determination image-region of mark object images matching, and exist and target pair in the predeterminable area of the previous frame image of current frame image
As images match latent image region when, obtain the coordinate position of the latent image region in the previous frame image;
Wherein, described is the image-region for meeting multiple matching conditions with destination object images match determination image-region, it is described with
The latent image region of destination object images match is the image-region for only meeting one of the plurality of matching condition;
The detector unit, specifically for the first setting area comprising the coordinate position for obtaining in the current frame image
Detect whether in domain exist and the destination object images match determination image-region, and with the presence or absence of with the target pair
As the latent image region of images match;
The determining unit, specifically for working as the determination existed in first setting regions with the destination object images match
During image-region, the determination image existed in the predeterminable area of the current frame image with the destination object images match is determined
Region;When there is the latent image region with the destination object images match in first setting regions, it is determined that described
There is the latent image region with the destination object images match in the predeterminable area of current frame image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210463919.5A CN103810696B (en) | 2012-11-15 | 2012-11-15 | Method for detecting image of target object and device thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210463919.5A CN103810696B (en) | 2012-11-15 | 2012-11-15 | Method for detecting image of target object and device thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103810696A CN103810696A (en) | 2014-05-21 |
CN103810696B true CN103810696B (en) | 2017-03-22 |
Family
ID=50707417
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210463919.5A Active CN103810696B (en) | 2012-11-15 | 2012-11-15 | Method for detecting image of target object and device thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103810696B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI550440B (en) * | 2014-12-09 | 2016-09-21 | 由田新技股份有限公司 | Method and system for detecting person to use handheld apparatus |
CN107920223B (en) * | 2016-10-08 | 2020-08-28 | 杭州海康威视数字技术股份有限公司 | Object behavior detection method and device |
CN108109132B (en) | 2016-11-25 | 2021-02-05 | 杭州海康威视数字技术股份有限公司 | Image analysis method and device |
CN106686308B (en) * | 2016-12-28 | 2018-02-16 | 平安科技(深圳)有限公司 | Image focal length detection method and device |
CN108388879B (en) * | 2018-03-15 | 2022-04-15 | 斑马网络技术有限公司 | Target detection method, device and storage medium |
CN111723603A (en) * | 2019-03-19 | 2020-09-29 | 杭州海康威视数字技术股份有限公司 | Material monitoring method, system and device |
CN113066121A (en) * | 2019-12-31 | 2021-07-02 | 深圳迈瑞生物医疗电子股份有限公司 | Image analysis system and method for identifying repeat cells |
CN111898581B (en) * | 2020-08-12 | 2024-05-17 | 成都佳华物链云科技有限公司 | Animal detection method, apparatus, electronic device, and readable storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1912950A (en) * | 2006-08-25 | 2007-02-14 | 浙江工业大学 | Device for monitoring vehicle breaking regulation based on all-position visual sensor |
CN101305616A (en) * | 2005-09-09 | 2008-11-12 | 索尼株式会社 | Image processing device and method, program, and recording medium |
CN101673403A (en) * | 2009-10-10 | 2010-03-17 | 安防制造(中国)有限公司 | Target following method in complex interference scene |
CN102087747A (en) * | 2011-01-05 | 2011-06-08 | 西南交通大学 | Object tracking method based on simplex method |
CN102567705A (en) * | 2010-12-23 | 2012-07-11 | 北京邮电大学 | Method for detecting and tracking night running vehicle |
-
2012
- 2012-11-15 CN CN201210463919.5A patent/CN103810696B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101305616A (en) * | 2005-09-09 | 2008-11-12 | 索尼株式会社 | Image processing device and method, program, and recording medium |
CN1912950A (en) * | 2006-08-25 | 2007-02-14 | 浙江工业大学 | Device for monitoring vehicle breaking regulation based on all-position visual sensor |
CN101673403A (en) * | 2009-10-10 | 2010-03-17 | 安防制造(中国)有限公司 | Target following method in complex interference scene |
CN102567705A (en) * | 2010-12-23 | 2012-07-11 | 北京邮电大学 | Method for detecting and tracking night running vehicle |
CN102087747A (en) * | 2011-01-05 | 2011-06-08 | 西南交通大学 | Object tracking method based on simplex method |
Also Published As
Publication number | Publication date |
---|---|
CN103810696A (en) | 2014-05-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103810696B (en) | Method for detecting image of target object and device thereof | |
US11010905B2 (en) | Efficient object detection and tracking | |
CN102750709B (en) | Video is utilized to detect the method and apparatus of behavior of fighting | |
KR20190046351A (en) | Method and Apparatus for Detecting Intruder | |
CN108875531B (en) | Face detection method, device and system and computer storage medium | |
CN104680145B (en) | The on off state change detecting method and device of a kind of | |
CN112597877A (en) | Factory personnel abnormal behavior detection method based on deep learning | |
CN111914656A (en) | Personnel behavior detection method and device, electronic equipment and storage medium | |
CN112580434B (en) | Face false detection optimization method and system based on depth camera and face detection equipment | |
CN109544516A (en) | Image detecting method and device | |
CN114663390A (en) | Intelligent anti-pinch method, device, equipment and storage medium for automatic door | |
CN114882251A (en) | Internet of things big data intelligent video security monitoring method and device | |
CN110865415A (en) | Security inspection method and device | |
CN110580708B (en) | Rapid movement detection method and device and electronic equipment | |
CN112597903B (en) | Electric power personnel safety state intelligent identification method and medium based on stride measurement | |
CN108446581A (en) | A kind of unmanned plane detection method in adverse circumstances | |
CN114764895A (en) | Abnormal behavior detection device and method | |
CN111695404B (en) | Pedestrian falling detection method and device, electronic equipment and storage medium | |
CN116758493A (en) | Tunnel construction monitoring method and device based on image processing and readable storage medium | |
CN112435240B (en) | Deep vision mobile phone detection system for workers to illegally use mobile phones | |
CN112861711A (en) | Regional intrusion detection method and device, electronic equipment and storage medium | |
KR20140143918A (en) | Method and Apparatus for Detecting Foregroud Image with Separating Foregroud and Background in Image | |
CN109740518A (en) | The determination method and device of object in a kind of video | |
CN114979567B (en) | Object and region interaction method and system applied to video intelligent monitoring | |
CN106595604A (en) | Nuclear power plant perimeter low-altitude parabolic behavior detection method and device |
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 |