CN102131050A - Method for recognizing multi-object image - Google Patents
Method for recognizing multi-object image Download PDFInfo
- Publication number
- CN102131050A CN102131050A CN 201010188508 CN201010188508A CN102131050A CN 102131050 A CN102131050 A CN 102131050A CN 201010188508 CN201010188508 CN 201010188508 CN 201010188508 A CN201010188508 A CN 201010188508A CN 102131050 A CN102131050 A CN 102131050A
- Authority
- CN
- China
- Prior art keywords
- image
- image section
- pixel
- described image
- recognition method
- 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
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a method for recognizing a multi-object image, which is used for recognizing a plurality of objects in the image. Each object consists of a plurality of image segments. The method comprises the following steps of: capturing each pixel of the image in sequence; determining an image segment initial point of an unknown object; recording information of the image segment point by point from the image segment initial point; determining an image segment terminal point of the unknown object; distinguishing the object to which the image segment belongs according to the image segment initial point and the image segment terminal point; and judging an invalid or combined object to release a data space of the object.
Description
Technical field
The present invention relates to a kind of image recognition methods, particularly a kind ofly can handle and need not write down the method that total image can carry out the multi-object image identification immediately.
Background technology
Present image processing technique, object (Objects) for identification any amount in an image, often need to cooperate different image identification algorithms, and general algorithm is along with the increase of desiring identification objects quantity in the described image, it is complicated that the calculating of its algorithm also more becomes.For example must use complicated region growing (Region growing) algorithm, also therefore must in advance the image buffer (Image buffer) of whole image (total image) in image processing system be stored, after waiting to collect the information of all images, just can carry out complicated recognizer to pick out each object in the described image.Thus, not only take the memory resource of image buffer during identification, also very expend time in.
Summary of the invention
Therefore, a purpose of the present invention is to provide a kind of method that total image can carry out the multi-object image identification that need not write down, and it can effectively save memory resource.
Another purpose of the present invention is to provide a kind of and has extendibility, is not subjected to the number of objects restriction in the image and can immediately pick out the multi-object image recognition method of each object.
Still a further object of the present invention is to provide a kind of multi-object image recognition method, and whether it can identification have object to be merged or be invalid to determine whether discharge the hash space in identification objects, uses the shared memory resource of further saving.
For reaching above-mentioned purpose, multi-object image recognition method of the present invention is in order to a plurality of objects in identification one image, and wherein each object is made up of a plurality of image sections.Described method comprises the following steps: to capture in regular turn each pixel of described image; When acquisition one pixel, judge whether described pixel is an image section starting point of an image section of a unknown object; Write down the information of described image section from the pointwise of described image section starting point; When the information of the described image section of record, judge an image area segment endpoint of described unknown object; Differentiate the affiliated object of described image section according to described image section starting point and described image area segment endpoint; Judged whether that object finishes; When object finishes, judge whether end object is invalid; And, discharge the data space of invalidated object when described end object when being invalid.
The present invention provides a kind of multi-object image recognition method in order to a plurality of objects in identification one image in addition, and wherein each object is made up of a plurality of image sections.Described method comprises the following steps: to capture in regular turn each pixel of described image; When acquisition one pixel, judge whether described pixel is an image section starting point of an image section of a unknown object; Write down the information of described image section from the pointwise of described image section starting point; When the information of the described image section of record, judge an image area segment endpoint of described unknown object; Differentiate the affiliated object of described image section according to described image section starting point and described image area segment endpoint; Judged whether that object is merged; And merged as object, discharge by the data space of combining objects.
The present invention provides a kind of multi-object image recognition method in order to a plurality of objects in identification one image in addition, and wherein each object is made up of a plurality of image sections.Described method comprises the following steps: to capture in regular turn each pixel of described image; When acquisition one pixel, judge whether this pixel is an image section starting point of an image section of a unknown object; Write down the information of described image section from the pointwise of described image section starting point; When the information of the described image section of record, judge an image area segment endpoint of described unknown object; Utilize the spatial coherence of each object image section in described image section and adjacent lines or the adjacent column to differentiate the affiliated object of described image section; And judge invalid or by combining objects not discharge its data space.
In another form of implementation, the information that multi-object image recognition method of the present invention comprises the following steps: to compile the described image section that writes down in addition object under it, the information of wherein said image section (for example end points coordinate and pixel gray level value) is recorded in the register.
In the multi-object image recognition method of the present invention, when delegation or a row pixel had been captured, the object of no coherent video section then was judged as end in described row or described row; When the size of end object or mean flow rate are not in a preset range, judge that then described end object is invalid; Then judging when described image section belongs to a plurality of object simultaneously has object to be merged.
Description of drawings
Fig. 1 has shown the calcspar of the image processing system of the embodiment of the invention.
Fig. 2 has shown the schematic diagram in order to the image that multi-object image recognition method of the present invention is described.
Fig. 3 has shown the flow chart of the multi-object image recognition method of the embodiment of the invention.
Fig. 4 has shown the schematic diagram of judging the object merging in the multi-object image recognition method of the present invention.
Fig. 5 has shown the schematic diagram of judging the object end in the multi-object image recognition method of the present invention.
Fig. 6 has shown the schematic diagram of judging object invalidation in the multi-object image recognition method of the present invention.
The main element symbol description
1,4,5,6 images, 11 circular object
12 triangle object, 111～114,121～125 image sections
101～109 step 3 image processing systems
31 image sensor, 311 pixels
32 analog-digital converters, 33 image process units
34 registers 401,402,501, a, b object
X
0～X
4One-row pixels
111a, 112a, 121a, 122a image section starting point
111b, 112b, 121b, 122b image area segment endpoint
Embodiment
For allow above-mentioned and other purposes of the present invention, feature and advantage can be more obvious, hereinafter will cooperate appended diagram, be described in detail below.
As shown in Figure 1, the present invention need not write down the embodiment that total image can carry out the method for multi-object image identification and is applied in the image processing system 3.Described image processing system 3 has image sensor (Image sensor) 31, analog-digital converter (A/D Converter) 32, image process unit (Imageprocessor) 33 and register (Register) 34.Described image sensor 31 can be made by CCD or cmos element, in order to respond to the object reverberation of (figure does not show), launch light or to cover shadow to form the analog image signal.Then, the analog image signal is transferred into analog-digital converter 32 and is converted to digital image signal.33 computings of being responsible for most of signal of video signal of described image process unit.Among other embodiment, described analog-digital converter 32 can be included in the described image sensor 31.
Mandatory declaration be that the image processing system 3 of present embodiment can be used for the making a video recording discriminating function of image-taking devices such as making video recording maybe can be to realize by a video-unit pick-up image and in the mode that the recognition software that is installed in computer is carried out the discriminating function of described image.In addition, because the aufbauprinciple of image sensor 31, analog-digital converter 32, image process unit 33 and other related elements is a known technology, so do not giving unnecessary details.Main concept of the present invention is to cooperate register 34 to carry out the discriminating function of a plurality of objects in the image with image process unit 33, therefore following will only being described with regard to the part that is relevant to the principle of the invention.In addition, among other embodiment, described register 34 also can be included in described image process unit 33 or other members, is not limited to the content that Fig. 1 discloses.
Cooperating Figure 1 and Figure 2, the present invention need not write down method that total image can carry out the multi-object image identification is that the object of any amount that had for the image 1 (as Fig. 2) that image sensor 31 is captured carries out identification.In the present embodiment, in the image 1 to be identified to as if be the step that example explanation is carried out identification with a circular object 11 and a triangle object 12.
Mandatory declaration be, because image sensor 31 has a plurality of determinant inducing pixels (pixel) 311, and described a plurality of pixel 311 is to respond to each object 11,12 line by line or by the mode of row, therefore, an object of herein image sensor 31 being sensed is called an image section (Image Segment) at the partial image of each row or each row.Circular object 11 for example shown in Figure 2 has four lines image section 111～114, and triangle object 12 has five-element's image section 121～125, by that analogy.Be understandable that the image section is defined as one-row pixels or a row pixel can determine according to the running of described image sensor 31.Though in the aftermentioned explanation image section is illustrated that with row it only is exemplary, be not in order to limit the present invention.
Cooperation Fig. 1～shown in Figure 3, the present invention need not write down total image can be carried out each step of method of multi-object image identification and action principle details are as follows: at first, described image sensor 31 acquisitions one image 1, described image process unit 33 captures each pixel value (step 101) of described image 1 in regular turn from described analog-digital converter 32.Among a kind of embodiment, described image process unit 33 constantly reads each pixel value of every row by that analogy from first capable each pixel value that begins from left to right to read in the described row.
When reading a pixel, judge whether described pixel is an image section starting point of an image section of a unknown object; If then its coordinate and GTG value are stored to register 34 (step 102).
Then, begin from described image section starting point, pointwise is write down the information (for example, GTG value) of described image section and is stored to register 34 (step 103).
In the information of the described image section of record, judge an image area segment endpoint of described unknown object, and its coordinate and GTG value are stored to register 34 (step 104); Wherein step 102 judges whether that to 104 the mode of object image information occurs greater than the pixel value of a systemic presupposition threshold value (threshold) for judging whether; The setting means of wherein said predetermined threshold value for example, can decide in its sole discretion by the user, can according to one with reference to image set or according to the variable quantity of contiguous pixels GTG value as described threshold value, for example can be according to a ratio of the mean flow rate of described image sensor 31 last the image that capture or a biasing (offset) as described predetermined threshold value, but predetermined threshold value of the present invention is not limited to this.
Then, utilize the spatial coherence of each object image section in the image section starting point of image section of described unknown object and image area segment endpoint and adjacent lines or the adjacent column to differentiate the affiliated object (step 105) of described image section.
In the present embodiment, when described spatial coherence meets following formula 1, can judge that then described unknown object image section belongs to object i:
Seg_L≤Preline_Obji_R; And
Seg_R≥Preline_Obji_L。
Wherein, when supposing to read to the y line data in the image 1, then Preline_Obji_R represents the right-hand terminal point X coordinate figure of image section of each described object i of the capable appearance of y-1; Preline_Obji_L represents the image section left starting point X coordinate figure of each described object i of the capable appearance of y-1; Seg_L represents to read the image section left starting point X coordinate figure of the unknown object of the capable appearance of y; Seg_R represents to read the right-hand terminal point X of the image section coordinate figure of the unknown object of the capable appearance of y.That is, judge whether described image section starting point is arranged in (for example right side) after the starting point whether (for example left side) before the terminal point of adjacent lines or adjacent column one object image area section and described image area segment endpoint be arranged in adjacent lines or adjacent column one object image area section.
The information of compiling the described image section that writes down under it object (step 106) and be stored to described register 34.
In order to reduce taking up room of described register 34, judged whether that then object finishes or merged (step 107).When object is merged, then discharge by the shared data space of combining objects (step 108a), wherein a pair of as if be not merged into the image section of judging a unknown object and whether belong to a plurality of objects simultaneously.When object finishes, judge whether described object is invalid (step 108b), wherein a pair of as if not finish be that the object of no coherent video section in described row or described row (promptly an object is not compiled the image section of any unknown object) then is judged as end when row or one-row pixels acquisition finish.When described object is invalid, then discharge the shared data space (step 109) of described invalidated object, wherein judge the whether invalid mode of object for size, overall brightness or the mean flow rate of judging object whether in a preset range.Undelivered object information then continues to be stored in the described register 34, then, carries out the judgement of (if existence) of next image section in this journey with same steps as again; After having captured described image 1 all pixel values, also finish the multi-object identification of described image 1 simultaneously.
Cooperation Fig. 1～shown in Figure 3 then illustrates multi-object image recognition method of the present invention.Suppose in image 1, to read each pixel line by line from the 1st row beginning, because coordinate (1,3) locate to occur greater than the pixel value of systemic presupposition threshold value, therefore just write down the coordinate figure of starting point 111a of object 11 and GTG value in register 34, the then information of pointwise recording image section 111 and being stored in the register 34, up to the terminal point 111b that runs into described image section 111, write down the coordinate figure of described terminal point 111b and GTG value again in register 34.Yet owing to then have the information of another image section 121 in first row, the information that the coordinate figure that also need store the starting point 121a of described image section 121 and terminal point 121b again and GTG value and pointwise are write down is in register 34.Therefore, 34 accumulative totals of described register this moment have the image section 111 and 121 of two objects.
Then, described image processing unit 33 receives the image information of second row of described image 1 in regular turn, and write down left starting point coordinate value 112a, 122a and the GTG value thereof of each unknown object image section that this row occurs, and right-hand terminal point coordinate value 112b, 122b and the GTG value thereof of each unknown object image section, and when reading to the right-hand terminal point of each unknown object image section, whether belong to each object 11,12 as criterion to distinguish with formula 1 immediately, with order right by a left side, from top to bottom is principle, writes down union line by line to last column.Therefore, when having captured the value of described image 1 all pixels 311, can immediately pick out each the described object 11,12 in the described image fully.
Conclude above-mentionedly, the present invention need not write down the method that total image can carry out the multi-object image identification and have following advantage:
1. image recognition method of the present invention carries out image identification immediately in the mode of depositing, and need not to write down whole image, therefore can save memory resource.
2. image recognition method of the present invention is not limited by the number of objects in the image because algorithm is simple, therefore any amount of object all can be come out by identification, not only has extendibility, also can immediately pick out each object.
In addition, in order further to reduce the use amount of register 34 in the process identification process, when running into, object under the register 34 stored image sections merged or during the information conforms of end object is invalid condition, the space that stores described object in the register 34 can be discharged, reduce the use amount of described register 34 by this.
Please refer to shown in Figure 4ly, it has shown in the multi-object image recognition method of the present invention the schematic diagram of judging that object merges.When described image process unit 33 is handled X
0During the row pixel, store two objects 401 and 402 in the described register 34; Yet when described image process unit 33 is handled to X
1During the row pixel, recognize described object 401 and 402 and belong to same target (because X
1The image section of row belongs to object 401 and 402 simultaneously), thereby described object 401 and 402 can be merged and discharges described object 401 or 402 shared data spaces (step 108a).
Please refer to shown in Figure 5ly, it has shown in the multi-object image recognition method of the present invention the schematic diagram of judging that object finishes.When described image process unit 33 is handled to X
2During the row pixel, because at X
2No any image section belongs to object 501 in the row pixel, therefore recognizes this object 501 and has finished, and judges then whether this object 501 is invalid (step 108b).
Please refer to shown in Figure 6ly, it has shown in the multi-object image recognition method of the present invention the schematic diagram of judging object invalidation.The mode of judging object invalidation among the present invention is for for example to judge that whether size, object overall brightness or the mean flow rate of object are between a preset range., object overall brightness too small when object size or mean flow rate are crossed when low, and it may be the noise in the described image 1, for example handle to X when described image process unit 33 among Fig. 6
3Row during pixel object a finished and undersized and be regarded as invalidated object, thereby can discharge the shared space of described object a.When, object overall brightness excessive when object size or mean flow rate were too high, it may be the image of other light sources or object in described image processing system 3 environment of living in, for example handled to X when described image process unit 33 among Fig. 6
4Row during pixel object b finished and oversize and be regarded as invalidated object, thereby can discharge the shared space of described object b, wherein the maximum of size can preestablish according to the object size of described image processing system 3 identification of answering.
Among the present invention, described register 34 can discharge the invalid data space that originally stored, and provides as subsequent step and use.Therefore except that the use amount that can reduce register 34, because the space of described register 34 is limited and fixing, therefore the release by the space can be in order to pick out how effective object.
The foregoing description only is used for exemplifying form of implementation of the present invention, and explains technical characterictic of the present invention, is not to be used for limiting protection category of the present invention.Any those skilled in the art can unlabored change or the arrangement of the isotropism scope that all belongs to the present invention and advocated, and the scope of the present invention should be as the criterion with the scope that claims were defined.
Claims (15)
1. multi-object image recognition method, in order to a plurality of objects in identification one image, each object is made up of a plurality of image sections, and described method comprises the following steps:
Capture each pixel of described image in regular turn;
When acquisition one pixel, judge whether this pixel is an image section starting point of an image section of a unknown object;
Write down the information of described image section from the pointwise of described image section starting point;
When the information of the described image section of record, judge an image area segment endpoint of described unknown object;
Differentiate the affiliated object of described image section according to described image section starting point and described image area segment endpoint;
Judged whether that object finishes;
When object finishes, judge whether end object is invalid; And
When described end object when being invalid, discharge the data space of invalidated object.
2. multi-object image recognition method according to claim 1, described method comprise the following steps: also to have judged whether that object is merged.
3. multi-object image recognition method according to claim 2, described method also comprises: when object is merged, discharge by the data space of combining objects.
4. multi-object image recognition method according to claim 1 wherein, judges whether whether size, overall brightness or the mean flow rate of end object invalid to judge described end object in a preset range.
5. multi-object image recognition method according to claim 1, described method also comprises: compile the affiliated object of the information of the described image section that is write down to this image section.
6. multi-object image recognition method according to claim 1 wherein, captures each row of described image or each pixel of each row in regular turn.
7. multi-object image recognition method according to claim 6, wherein, when delegation or a row pixel had been captured, the object of no coherent video section then was judged as end in described row or described row.
8. multi-object image recognition method, in order to a plurality of objects in identification one image, each object is made up of a plurality of image sections, and described method comprises the following steps:
Capture each pixel of described image in regular turn;
When acquisition one pixel, judge whether described pixel is an image section starting point of an image section of a unknown object;
Write down the information of described image section from the pointwise of described image section starting point;
When the information of the described image section of record, judge an image area segment endpoint of described unknown object;
Differentiate the affiliated object of described image section according to described image section starting point and described image area segment endpoint;
Judged whether that object is merged; And
When object is merged, discharge by the data space of combining objects.
9. multi-object image recognition method according to claim 8 wherein, judges whether described image section belongs to a plurality of objects simultaneously to have judged whether that object is merged.
10. multi-object image recognition method according to claim 8 wherein, captures each row of described image or each pixel of each row in regular turn.
11. multi-object image recognition method according to claim 8, described method also comprise the following steps: to compile the affiliated object of the information of the described image section that is write down to this image section.
12. a multi-object image recognition method, in order to a plurality of objects in identification one image, each object is made up of a plurality of image sections, and described method comprises the following steps:
Capture each pixel of described image in regular turn;
When acquisition one pixel, judge whether this pixel is an image section starting point of an image section of a unknown object;
Write down the information of described image section from the pointwise of described image section starting point;
When the information of the described image section of record, judge an image area segment endpoint of described unknown object;
Utilize the spatial coherence of each object image section in described image section and adjacent lines or the adjacent column, differentiate the affiliated object of described image section; And
Judge invalid or by combining objects to discharge its data space.
13. multi-object image recognition method according to claim 12 wherein, judges according to size, overall brightness or the mean flow rate of object whether described object is invalid.
14. multi-object image recognition method according to claim 12 wherein, judges whether described image section belongs to a plurality of objects simultaneously to have judged whether that object is merged.
15. multi-object image recognition method according to claim 12, wherein, described spatial coherence is for after judging whether described image section starting point is arranged in the preceding and described image area segment endpoint of the image area segment endpoint of adjacent lines or adjacent column one object and whether is positioned at the image section starting point of object described in adjacent lines or the adjacent column.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010188508 CN102131050A (en) | 2010-01-19 | 2010-05-31 | Method for recognizing multi-object image |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010001045.2 | 2010-01-19 | ||
| CN201010001045 | 2010-01-19 | ||
| CN 201010188508 CN102131050A (en) | 2010-01-19 | 2010-05-31 | Method for recognizing multi-object image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102131050A true CN102131050A (en) | 2011-07-20 |
Family
ID=44268924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010188508 Pending CN102131050A (en) | 2010-01-19 | 2010-05-31 | Method for recognizing multi-object image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102131050A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103970262A (en) * | 2013-02-06 | 2014-08-06 | 原相科技股份有限公司 | Optical pointing system |
| US9684386B2 (en) | 2013-01-29 | 2017-06-20 | Pixart Imaging Inc. | Optical pointing system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI234396B (en) * | 2004-05-10 | 2005-06-11 | Jau-Hung Jang | Image capturing/storage device to capture a specific targe image and method thereof |
| TW200639734A (en) * | 2005-05-02 | 2006-11-16 | Pixart Imaging Inc | Method for recognizing objects in an image without recording the image in its entirety |
| TW200639735A (en) * | 2005-05-02 | 2006-11-16 | Pixart Imaging Inc | Method for dynamically recognizing objects in an image based on diversities of object characteristics and system for using the same |
| CN101499169A (en) * | 2009-02-13 | 2009-08-05 | 北京中星微电子有限公司 | Method and apparatus for scanning connected region in combined image |
-
2010
- 2010-05-31 CN CN 201010188508 patent/CN102131050A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI234396B (en) * | 2004-05-10 | 2005-06-11 | Jau-Hung Jang | Image capturing/storage device to capture a specific targe image and method thereof |
| TW200537925A (en) * | 2004-05-10 | 2005-11-16 | zhao-hong Zhang | Image capturing/storage device to capture a specific targe image and method thereof |
| TW200639734A (en) * | 2005-05-02 | 2006-11-16 | Pixart Imaging Inc | Method for recognizing objects in an image without recording the image in its entirety |
| TW200639735A (en) * | 2005-05-02 | 2006-11-16 | Pixart Imaging Inc | Method for dynamically recognizing objects in an image based on diversities of object characteristics and system for using the same |
| CN101499169A (en) * | 2009-02-13 | 2009-08-05 | 北京中星微电子有限公司 | Method and apparatus for scanning connected region in combined image |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9684386B2 (en) | 2013-01-29 | 2017-06-20 | Pixart Imaging Inc. | Optical pointing system |
| US9766717B2 (en) | 2013-01-29 | 2017-09-19 | Pixart Imaging Inc. | Optical pointing system |
| US9958961B2 (en) | 2013-01-29 | 2018-05-01 | Pixart Imaging Inc. | Optical pointing system |
| US10228772B2 (en) | 2013-01-29 | 2019-03-12 | Pixart Imaging Inc. | Remote controller |
| CN103970262A (en) * | 2013-02-06 | 2014-08-06 | 原相科技股份有限公司 | Optical pointing system |
| CN103970262B (en) * | 2013-02-06 | 2018-01-16 | 原相科技股份有限公司 | Optical profile type pointing system |
| CN107992198A (en) * | 2013-02-06 | 2018-05-04 | 原相科技股份有限公司 | Optical profile type pointing system |
| CN107992198B (en) * | 2013-02-06 | 2021-01-05 | 原相科技股份有限公司 | Optical pointing system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI419061B (en) | Method for recognizing multiple objects | |
| US9053389B2 (en) | Hough transform for circles | |
| CN109147341B (en) | Violation vehicle detection method and device | |
| EP1271933A2 (en) | Bad pixel detection and correction in an image sensing device | |
| US20090097704A1 (en) | On-chip camera system for multiple object tracking and identification | |
| CN109241345B (en) | Video positioning method and device based on face recognition | |
| US10791252B2 (en) | Image monitoring device, image monitoring method, and recording medium | |
| JP2010526385A (en) | Sequential image alignment | |
| CN110572636B (en) | Camera contamination detection method and device, storage medium and electronic equipment | |
| Dugad et al. | Robust video shot change detection | |
| JP2010509651A (en) | Image recognition apparatus and method | |
| US20060245652A1 (en) | Method for recognizing objects in an image without recording the image in its entirety | |
| CN103279952A (en) | Target tracking method and device | |
| US20100026903A1 (en) | Motion vector detection device, motion vector detection method, and program | |
| US20210319572A1 (en) | Error reduction in depth maps | |
| US8699753B2 (en) | Object detecting device, image dividing device, integrated circuit, method of detecting object, object detecting program, and recording medium | |
| CN102131050A (en) | Method for recognizing multi-object image | |
| CN105744184B (en) | Bad pixel correction method and the device for using this method | |
| CN103458173A (en) | Device, method and terminal for processing image relevant data | |
| TW201421423A (en) | Image sensor and operating method thereof | |
| CN100468442C (en) | Method for performing multi-object image identification without recording full image | |
| CN116249015A (en) | Camera shielding detection method and device, camera equipment and storage medium | |
| JP2000348284A (en) | Device and method for vehicle detection and computer- readable recording medium where vehicle detecting program is recorded | |
| WO1998012869A1 (en) | Movement detecting apparatus | |
| JP5616743B2 (en) | Imaging apparatus and image processing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110720 |