CN101198033A - Locating method and device for foreground image in binary image - Google Patents
Locating method and device for foreground image in binary image Download PDFInfo
- Publication number
- CN101198033A CN101198033A CNA200710303785XA CN200710303785A CN101198033A CN 101198033 A CN101198033 A CN 101198033A CN A200710303785X A CNA200710303785X A CN A200710303785XA CN 200710303785 A CN200710303785 A CN 200710303785A CN 101198033 A CN101198033 A CN 101198033A
- Authority
- CN
- China
- Prior art keywords
- section
- sign
- zone
- image
- foreground
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000004458 analytical method Methods 0.000 claims description 21
- 230000011218 segmentation Effects 0.000 claims description 13
- 238000013316 zoning Methods 0.000 claims description 12
- 239000000284 extract Substances 0.000 claims description 10
- 230000004807 localization Effects 0.000 claims description 10
- 238000004891 communication Methods 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 abstract description 7
- 238000002054 transplantation Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 12
- 230000000694 effects Effects 0.000 description 11
- 238000012545 processing Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a locating method and a locating device for a foreground image in a binary image, comprising the following steps: the binary image is divided on the basis of regions and foreground pixels in various regions are obtained after scanning; continuous foreground pixels in the same region are organized into segments; the foreground image is obtained after incorporating communicated segments in various regions. When the invention is used in an intelligent video monitoring system and moving targets of images of each frame can be detected, timeliness of intelligent video monitoring is greatly improved and transplantation towards hardware is easy.
Description
Technical field
The present invention relates to the video analysis technology, the localization method and the device of foreground image in particularly a kind of bianry image.
Background technology
In intelligent video monitoring, it is the basis of most of intelligent video analysis function that the moving target in the video is positioned, and such as in utilizations such as people's flow analysis, vehicle analysis, moving body track, all needs the moving target in the video is positioned.Therefore the efficient of location algorithm directly has influence on the operational efficiency of video monitoring system.The method that moving target is positioned has a lot, and major part wherein is to utilize the method for image subtraction that the background parts in the video is removed, and remainder is obtained bianry image as prospect, bianry image is handled oriented moving target then.Because moving target is the zone of a connection, be exactly a kind of simple effective method based on the connected domain analysis of foreground pixel.Therefore, the efficient of obtaining the foreground image of bianry image becomes the key to the moving target location efficiency.
Prior art is generally carried out the processing of foreground image based on bianry image, the gray value of background area pixels is 0, and the gray value of each target area pixel is 1.Background can be regarded as piece image usually, is used for writing down the imaging of the object that remains static for a long time in the fixing three dimensions.About the generation and the renewal of background image, having proposed many methods can realize.When in this fixing three dimensions moving object being arranged, moving object will cover a part of background area when imaging, and these zones that are capped are exactly prospect, and common title prospect is the target area when target detection.Image-region except prospect then becomes the background area.Prior art has two kinds of comparison typical methods to handle.First kind is the algorithm of seed search, and this method is searched for bianry image from top to bottom, when finding a foreground pixel (being the seed point), just this pixel is put into storehouse.Search for the neighborhood of this pixel then,, just foreground pixel is put into storehouse if foreground pixel is arranged in the neighborhood.After the neighborhood of each pixel in the storehouse has all been searched for, just can form foreground image with all foreground pixel points that are communicated with seed point.The realization of this method is simple relatively, but its deficiency is: operational efficiency is very low, but also needs storehouse to deposit pixel, having relatively high expectations to memory space when connected domain is big.Second method is introduced the thought of scanning, from left to right, from top to bottom bianry image is scanned.And need the connectedness of the current pixel that just is being scanned of sign and several neighbour's pixels that before it, are scanned, have several different situations to consider respectively: if having only one in its neighborhood that has been scanned is 1, so it is designated and is attached thereto logical pixel, if it is connected with two or more targets, can think that then these targets are actually one.This algorithm needs twice scanning, identifies the connected relation of all pixels and neighborhood territory pixel for the first time, scans the pixel that will identify difference but be actually same connected region for the second time and is merged together.Second method does not need storehouse, only needs twice scanning just can finish the analysis of connected domain, so than the efficient height of first method.But its deficiency is: this method is based on also that pixel finishes, when image bigger, foreground pixel more for a long time, efficient can descend, that is, its efficient is subject to processing how many restrictions of the pixel of image.
Summary of the invention
The invention provides the localization method and the device of foreground image in a kind of bianry image, in order to improve the treatment effeciency of connected domain in binary Images Processing.
The invention provides the localization method of foreground image in a kind of bianry image, comprise the steps:
Foreground pixel in each zone is divided and obtained to bianry image after scanning by the zone;
With the foreground pixel section of being organized as continuous in the same area;
The section that is communicated with in each zone is merged the back obtain foreground image.
Preferably, the described section that is communicated with in each zone is merged is specially:
During with the foreground pixel section of being organized as continuous in the same area section is identified;
Be communicated with in the zones of different the section sign unify;
Section sign in the bianry image is same according to sign merges.
Preferably, the described division by the zone is specially: the ranks during with scanning are that unit carries out area dividing.
Preferably, described zone is the delegation or row of scanning.
Preferably, the described section that is communicated with in each zone is merged is specially:
To the line scanning of whenever advancing, the foreground pixel section of being organized as that every row is continuous, and section identified;
Described sign with the section that is communicated with in the zones of different is unified to comprise: each section that extracts is carried out the connected domain analysis, section carry out same sign with what communicate in the adjacent lines, the segment identification that does not communicate is inequality; To identifying difference but be labeled as and be connected between the section that is connected; According to dated communication information, the different section unifications that still belong to same connected region of sign are identical sign;
All sections of like-identified are merged.
The present invention also provides the positioner of foreground image in a kind of bianry image, comprising:
Acquisition module is used for bianry image is divided and obtained by the zone foreground pixel in each zone after scanning;
Segmentation module is used for the foreground pixel section of being organized as that the same area is continuous;
Merge module, be used for that the section that each zone is communicated with is merged the back and obtain foreground image.
Preferably, described merging module comprises:
Identify unit is used for when described segmentation module the foreground pixel section of being organized as that the same area is continuous section being identified;
The unified unit of sign, the sign that is used for section that zones of different is communicated with is unified;
Merge cells is used for merging according to the sign section that the bianry image sign is same.
Preferably, described acquisition module further comprises the zoning unit, and the ranks when being used for scanning are that unit carries out area dividing;
Acquisition module is used for obtaining the foreground pixel in each described zoning dividing elements zone after scanning.
Delegation or the unit of classifying as when preferably, described zoning unit is further used for scanning carry out area dividing.
Preferably, described segmentation module is further used for described acquisition module when whenever advancing line scanning, the foreground pixel section of being organized as that every row is continuous;
Identify unit is used for when described segmentation module the foreground pixel section of being organized as that every row is continuous section being identified;
Described sign is unified the unit and is comprised: analyze subelement, sign subelement, same subelement, wherein:
Analyze subelement, be used for each section that extracts carried out the connected domain analysis;
The sign subelement is used for according to the analysis of described analysis subelement the section that adjacent lines are communicated with being carried out same sign, and disconnected segment identification is inequality; To identifying difference but be labeled as and be connected between the section that is connected;
Same subelement is used for according to dated communication information, and the different section unifications that still belong to same connected region of sign are identical sign;
Merge cells is used for all sections of like-identified are merged.
Beneficial effect of the present invention is as follows:
In the location of the present invention's foreground image in bianry image, at first bianry image is divided and after scanning, is obtained the foreground pixel in each zone by the zone; Then with the foreground pixel section of being organized as continuous in the same area; Promptly the foreground pixel that at first utilizes scanning fast that each zone is connected each other is organized as " section ", then " section " that is interconnected is identified; According to sign the section that is communicated with in each zone is merged the back at last and obtain foreground image.Because continuous foreground pixel is organized into " section ", so concerning actual operation, no matter how many pixels has, when image becomes big, the quantity of pixel can increase, the quantity of " but section " does not increase, its existence is to the processing of the sign of " section " of same sign, and carry out the analysis of connected domain based on each pixel unlike existing method, when image becomes big, because of becoming, image must go to handle more pixel greatly when having the method computing now, operand also will increase greatly, causes the reduction of operation efficiency, because the present invention is not subjected to the restriction of pixel, so can improve treatment effeciency greatly to bianry image, especially obvious to big treatment of picture effect.
Description of drawings
Fig. 1 is the localization method implementing procedure schematic diagram of foreground image in the bianry image described in the embodiment of the invention;
Fig. 2 is for coming the foreground image localization method implementing procedure schematic diagram of zoning with row described in the embodiment of the invention;
Fig. 3 is that the connected domain based on section identifies schematic diagram described in the embodiment of the invention;
Fig. 4 is the positioning device structure schematic diagram of foreground image in the bianry image described in the embodiment of the invention;
Fig. 5 is the image actual effect schematic diagram after the binaryzation described in the embodiment of the invention;
Fig. 6 to Figure 11 is the actual effect schematic diagram after implementing described in the embodiment of the invention.
Embodiment
Core concept of the present invention is that the prospect with every row in the bianry image partly extracts, and the section of being organized as.Therefore the analysis of connected domain just can realize based on section, rather than traditional method based on pixel realizes.Advantage based on section is that efficiency of algorithm can not be subjected to the influence of image size.And since each section of every row separate, when identifying, only need to judge present segment and top delegation section between connected relation, improved efficiency of algorithm greatly.Describe below in conjunction with the embodiment of accompanying drawing the section of the present invention is based on.
Fig. 1 is the localization method implementing procedure schematic diagram of foreground image in the bianry image, as shown in the figure, can comprise the steps: in the foreground image location
Step 101, foreground pixel in each zone is divided and obtained to bianry image by the zone after scanning;
Step 102, with the foreground pixel section of being organized as continuous in the same area;
Step 103, section merging back that will be communicated with in each zone obtain foreground image.
Concrete, the described section that is communicated with in each zone is merged can be: during with the foreground pixel section of being organized as continuous in the same area section is identified; Be communicated with in the zones of different the section sign unify; According to sign the section that has same sign in the bianry image is merged.
Wherein, described by the zone divide can for: the ranks with when scanning are that unit carries out area dividing.When dividing, also can implement the ranks when not with scanning, such as marking particular area earlier as a zone, then the continuous foreground pixel point in this zone is organized by section again etc., and then merge by section, but consecutive hours can be because zone excessive or irregular and consuming time more judging foreground pixel like this, so best embody out efficient of the present invention when obviously, dividing by the ranks of scanning.
The location of the foreground image for behavior unit zoning the time implements to describe below, Fig. 2 is for coming the foreground image localization method implementing procedure schematic diagram of zoning with row, zone described in this example is the delegation or row of scanning, and as shown in the figure, can implement as follows this moment:
Step 201, input comprise the bianry image of moving target.
Step 202, from top to bottom is from left to right to the line scanning of whenever advancing.
Step 203, the foreground pixel section of being organized as that every row is continuous, and note.
Step 204, the connectedness of section and the sign of section are handled.
Below in conjunction with Fig. 3 this step is described, Fig. 3 is the connected domain sign schematic diagram based on section, and as shown in the figure, circle is represented foreground pixel, is the sign of this section in the circle; The rectangle frame S1 to S7 that contains circle represents 7 sections of this illustration meanings respectively, contain 4 continuous foreground pixels in every section, section is distributed in the four lines of scanning, wherein S1 is at first row, S2, S3 are at second row, and S4, S5, S6 are at the third line, and S7 is in fourth line, from top to bottom scan during scanning, promptly finish by the first row beginning fourth line.
Then, from top to bottom, from left to right each section that extracts carried out the connected domain analysis, be attached thereto logical section, so just be set to new logo, otherwise the sign of the section that just will be communicated with is with it composed to this section if this section can not find.If this section and multistage are connected, so just give this section, and the sign of the section of dated all connections belongs to same connected region with the minimum tax of sign in the section of all connections.
With Fig. 3 is example, when being scanned up to first section S1, because the section that the top of this section is not communicated with it, so this segment mark is designated as 1.When being scanned up to S2, the S1 of top is attached thereto logical, so just the mark of S1 is composed to S2.And S3 is not communicated with S2, and what therefore give S3 is new logo 2, so from top to bottom, from left to right scans successively.When being scanned up to S5, its top has two sections (S1 and S2) to be communicated with it, just compose that of mark minimum in these two sections to present segment this moment, the mark that is about to S5 is set to 1, indicate mark 1 and mark 2 simultaneously and belong to a connected region, and S6 is communicated with S3, thus the mark of S3 to compose to S6 be 2.
Step 205, the dated communication information of basis are labeled as identical sign with the different section unifications that still belong to same connected region of sign.
To belong to same zone in this step, but the different section of mark is merged into identical zone.1 just need merge into identical mark as being labeled as among Fig. 3,, also indicate simultaneously mark 1 and mark 2 and belong to a connected region, so be designated 1,2 period be communicated with in this example because be set to 1 at the mark of S5 with 2 section.When determining that it is same connected region, just S1 to S7 can be carried out unified sign so that next-step operation.
Step 206, with like-identified all the section collect, obtain the foreground target of all connections.
That is, it is at first to the line scanning of whenever advancing that the section that is communicated with in each zone is merged in this example enforcement, the foreground pixel section of being organized as that every row is continuous, and section identified; Then each section that extracts carried out the connected domain analysis, the section that communicates in the adjacent lines is carried out same sign, the segment identification that does not communicate is inequality; To identifying difference but be labeled as and be connected between the section that is connected; According to dated communication information, the different section unifications that still belong to same connected region of sign are identical sign; All sections with like-identified merge at last.
The present invention also provides the positioner of foreground image in a kind of bianry image, and in conjunction with above-mentioned explanation to localization method, the embodiment to device describes below.
Fig. 4 is the positioning device structure schematic diagram of foreground image in the bianry image, as shown in the figure, can comprise acquisition module 401, segmentation module 402 in the device, merge module 403; Also show bianry image among the figure, wherein circle is represented foreground pixel, and box indicating is to the section that foreground pixel divided, in device:
Acquisition module 401 is used for bianry image is divided and obtained by the zone pixel in each zone after scanning;
402 of segmentation modules are used for the foreground pixel section of being organized as that the same area is continuous;
Merging module 403 is used for the section that each zone is communicated with is merged back acquisition foreground image.
Acquisition module can further include the zoning unit in the device, and the ranks when being used for scanning are that unit carries out area dividing;
Acquisition module then obtains the foreground pixel in each described zoning dividing elements zone after scanning.
In the enforcement, the zoning unit can carry out area dividing with the delegation or the unit of classifying as in when scanning.
Merging module 403 in the device can comprise identify unit, the unified unit of sign, merge cells, wherein:
Identify unit is used for when described segmentation module the foreground pixel section of being organized as that the same area is continuous section being identified;
The unified unit of sign, the sign that is used for section that zones of different is communicated with is unified;
Merge cells is used for merging according to the sign section that the bianry image sign is same.
In the enforcement, when described acquisition module 401 when whenever advancing line scanning, 402 of described segmentation modules are used for the foreground pixel section of being organized as that every row is continuous; Merge identify unit in the module 403 when described segmentation module 402 the pixel organization section of being that every row is continuous, section is identified;
Sign in the merging module 403 is unified the unit and then can be comprised: analyze subelement, sign subelement, same subelement, wherein:
Analyze subelement, be used for each section that extracts carried out the connected domain analysis;
The sign subelement is used for according to the analysis of described analysis subelement the section that adjacent lines are communicated with being carried out same sign, and disconnected segment identification is inequality; To identifying difference but be labeled as and be connected between the section that is connected;
Same subelement is used for according to dated communication information, and the different section unifications that still belong to same connected region of sign are identical sign;
Merge at last in the module 403 merge cells with like-identified all the section merge.
As can be seen from the above-described embodiment, the present invention proposes moving target method for quick and the device that in intelligent video monitoring, can often use.The present invention finishes on bianry image, and bianry image is to subtract each other with current frame image and background image to obtain.Different with traditional connected domain algorithm based on pixel, the algorithm that the present invention proposes is based on that " section " realize, promptly at first utilize scanning fast that the foreground pixel that connects each other of every row is organized as " section ", then " section " that is interconnected identified.This algorithm is especially obvious to the effect of big image, because the foreground pixel point of big image is more, must cause tradition to increase based on the algorithm operation quantity of single pixel, but image increases the number that does not increase every row contact " section ".That is to say, after pixel groups is made into " section ", owing to continuous pixel groups will be made into " section ", so concerning actual operation, no matter how many pixels has, when image becomes big, the quantity of pixel can increase, the quantity of " but section " does not increase, its existence is to the processing of the sign of " section " of same sign, and in the prior art, traditional method is carried out the analysis of connected domain based on pixel, need handle each pixel, so when image becomes big, must go to handle more pixel greatly because of image becomes during its computing, operand also will increase greatly, so the method that the present invention proposes also can access higher running efficiency on to big treatment of picture.
With real processing results effect of the present invention is described again below.
Fig. 5 is the image actual effect schematic diagram after the binaryzation, Fig. 6 to Figure 11 is the actual effect schematic diagram after implementing, white edge among Fig. 6 to Figure 11 and white dashed line are the effects with the actual generation of the present invention, the white rectangle frame is the moving target that extracts according to the present invention, and movement locus is represented with white dashed line.Fig. 6 to Figure 11 shows is the effect signal of original video image, and the real goal that the present invention handles is actually the wherein image as shown in Figure 5 after the binaryzation, and white edge is the minimum rectangle that comprises white foreground target zone among the last figure.For better observing effect, white portion is added to obtains the effect of Fig. 6 to Figure 11 on the original image.When having carried out tracking, just can obtain to represent the white dashed line of pursuit path to each moving target.
Can find out obviously that from Fig. 6 to Figure 11 the present invention can well extract the moving target of all connections in publishing picture.In the reality,, improved the location efficiency of moving target greatly owing to proposed the connected component labeling processing method of the section of the present invention is based on.When the bianry image of 100 320*240 was tested by prior art, be average every running time: 0.949495ms.And with being the running time of average every of method of the present invention: 0.161616ms.To big image, the difference of operational efficiency is then more obvious.Obviously, because the raising of treatment effeciency, in intelligent video monitoring system, when the image of every frame is all needed to carry out motion target detection, can improve the real-time of intelligent video monitoring greatly undoubtedly, and be easy to transplant toward hardware.
Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.
Claims (10)
1. the localization method of foreground image in the bianry image is characterized in that, comprises the steps:
Foreground pixel in each zone is divided and obtained to bianry image after scanning by the zone;
With the foreground pixel section of being organized as continuous in the same area;
The section that is communicated with in each zone is merged the back obtain foreground image.
2. the method for claim 1 is characterized in that, the described section that is communicated with in each zone is merged is specially:
During with the foreground pixel section of being organized as continuous in the same area section is identified;
Be communicated with in the zones of different the section sign unify;
Section sign in the bianry image is same according to sign merges.
3. method as claimed in claim 1 or 2 is characterized in that, the described division by the zone is specially: the ranks during with scanning are that unit carries out area dividing.
4. method as claimed in claim 3 is characterized in that, described zone is the delegation or row of scanning.
5. method as claimed in claim 4 is characterized in that, the described section that is communicated with in each zone is merged is specially:
To the line scanning of whenever advancing, the foreground pixel section of being organized as that every row is continuous, and section identified;
Described sign with the section that is communicated with in the zones of different is unified to comprise: each section that extracts is carried out the connected domain analysis, section carry out same sign with what be communicated with in the adjacent lines, disconnected segment identification is inequality; To identifying difference but be labeled as and be connected between the section that is connected; According to dated communication information, the different section unifications that still belong to same connected region of sign are identical sign;
All sections of like-identified are merged.
6. the positioner of foreground image in the bianry image is characterized in that, comprising:
Acquisition module is used for bianry image is divided and obtained by the zone foreground pixel in each zone after scanning;
Segmentation module is used for the foreground pixel section of being organized as that the same area is continuous;
Merge module, be used for that the section that each zone is communicated with is merged the back and obtain foreground image.
7. device as claimed in claim 6 is characterized in that, described merging module comprises:
Identify unit is used for when described segmentation module the foreground pixel section of being organized as that the same area is continuous section being identified;
The unified unit of sign, the sign that is used for section that zones of different is communicated with is unified;
Merge cells is used for merging according to the sign section that the bianry image sign is same.
8. device as claimed in claim 7 is characterized in that described acquisition module further comprises the zoning unit, and the ranks when being used for scanning are that unit carries out area dividing;
Acquisition module is used for obtaining the foreground pixel in each described zoning dividing elements zone after scanning.
9. device as claimed in claim 8 is characterized in that, delegation or the unit of classifying as when described zoning unit is further used for scanning carry out area dividing.
10. device as claimed in claim 9 is characterized in that, described segmentation module is further used for described acquisition module when whenever advancing line scanning, the foreground pixel section of being organized as that every row is continuous;
Identify unit is used for when described segmentation module the foreground pixel section of being organized as that every row is continuous section being identified;
Described sign is unified the unit and is comprised: analyze subelement, sign subelement, same subelement, wherein:
Analyze subelement, be used for each section that extracts carried out the connected domain analysis;
The sign subelement is used for according to the analysis of described analysis subelement the section that adjacent lines are communicated with being carried out same sign, and disconnected segment identification is inequality; To identifying difference but be labeled as and be connected between the section that is connected;
Same subelement is used for according to dated communication information, and the different section unifications that still belong to same connected region of sign are identical sign;
Merge cells is used for all sections of like-identified are merged.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200710303785XA CN101198033A (en) | 2007-12-21 | 2007-12-21 | Locating method and device for foreground image in binary image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA200710303785XA CN101198033A (en) | 2007-12-21 | 2007-12-21 | Locating method and device for foreground image in binary image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101198033A true CN101198033A (en) | 2008-06-11 |
Family
ID=39548128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA200710303785XA Pending CN101198033A (en) | 2007-12-21 | 2007-12-21 | Locating method and device for foreground image in binary image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101198033A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299274B (en) * | 2008-06-18 | 2010-06-09 | 北京中星微电子有限公司 | Detecting method and system for moving fixed target |
| CN101763499A (en) * | 2008-12-25 | 2010-06-30 | 义隆电子股份有限公司 | Identification method for target image |
| CN101673401B (en) * | 2008-09-08 | 2013-03-27 | 索尼株式会社 | Image processing apparatus, method, and program |
| CN107240101A (en) * | 2017-04-13 | 2017-10-10 | 桂林优利特医疗电子有限公司 | Target area detection method and device, image partition method and device |
| WO2018058573A1 (en) * | 2016-09-30 | 2018-04-05 | 富士通株式会社 | Object detection method, object detection apparatus and electronic device |
| CN109300165A (en) * | 2018-09-14 | 2019-02-01 | 南京邮电大学 | A kind of novel target tracking localization method based on pixel characteristic |
-
2007
- 2007-12-21 CN CNA200710303785XA patent/CN101198033A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101299274B (en) * | 2008-06-18 | 2010-06-09 | 北京中星微电子有限公司 | Detecting method and system for moving fixed target |
| CN101673401B (en) * | 2008-09-08 | 2013-03-27 | 索尼株式会社 | Image processing apparatus, method, and program |
| CN101763499A (en) * | 2008-12-25 | 2010-06-30 | 义隆电子股份有限公司 | Identification method for target image |
| WO2018058573A1 (en) * | 2016-09-30 | 2018-04-05 | 富士通株式会社 | Object detection method, object detection apparatus and electronic device |
| CN109479118A (en) * | 2016-09-30 | 2019-03-15 | 富士通株式会社 | Method for checking object, object test equipment and electronic equipment |
| CN107240101A (en) * | 2017-04-13 | 2017-10-10 | 桂林优利特医疗电子有限公司 | Target area detection method and device, image partition method and device |
| CN109300165A (en) * | 2018-09-14 | 2019-02-01 | 南京邮电大学 | A kind of novel target tracking localization method based on pixel characteristic |
| CN109300165B (en) * | 2018-09-14 | 2022-08-30 | 南京邮电大学 | Novel target tracking and positioning method based on pixel characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | An edge embedded marker-based watershed algorithm for high spatial resolution remote sensing image segmentation | |
| CN101198033A (en) | Locating method and device for foreground image in binary image | |
| CN101499169B (en) | Method and apparatus for scanning connected region in combined image | |
| CN101799968B (en) | Detection method and device for oil well intrusion based on video image intelligent analysis | |
| CN105868708A (en) | Image object identifying method and apparatus | |
| US7720281B2 (en) | Visual characteristics-based news anchorperson segment detection method | |
| CN103646257A (en) | Video monitoring image-based pedestrian detecting and counting method | |
| CN104715251A (en) | Salient object detection method based on histogram linear fitting | |
| CN102842036A (en) | Intelligent multi-target detection method facing ship lock video monitoring | |
| CN106815587A (en) | Image processing method and device | |
| CN104252629A (en) | Target Detection And Tracking Method And System | |
| CN105354570A (en) | Method and system for precisely locating left and right boundaries of license plate | |
| US9747511B2 (en) | Image recognition device, image recognition method, program, and recording medium | |
| US8437503B2 (en) | Method for detecting a moving object in an image stream | |
| CN103092929B (en) | A kind of generation method and device of video frequency abstract | |
| Jin et al. | A color image segmentation method based on improved K-means clustering algorithm | |
| CN102073868A (en) | Digital image closed contour chain-based image area identification method | |
| Li | Fast multi-level connected component labeling for large-scale images | |
| Zhao et al. | A license plate recognition algorithm under low illumination environment | |
| Angeline et al. | Tracking and localisation of moving vehicle license plate via Signature Analysis | |
| Jianyong et al. | An edge-based approach for video text extraction | |
| He et al. | An algorithm for calculating objects’ shape features in binary images | |
| CN112330611A (en) | FPGA-oriented high-parallelism light spot segmentation method | |
| Zhang et al. | Marker-based watershed segmentation embedded with edge information | |
| Dhole et al. | Object Detection Using Blob Extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20080611 |