CN101715050B - Method and system for detecting dead pixels of image sensor - Google Patents
Method and system for detecting dead pixels of image sensor Download PDFInfo
- Publication number
- CN101715050B CN101715050B CN200910238438.2A CN200910238438A CN101715050B CN 101715050 B CN101715050 B CN 101715050B CN 200910238438 A CN200910238438 A CN 200910238438A CN 101715050 B CN101715050 B CN 101715050B
- Authority
- CN
- China
- Prior art keywords
- component value
- pixel
- bad point
- sequence
- monochrome
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 239000003086 colorant Substances 0.000 claims abstract description 24
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 24
- 238000012545 processing Methods 0.000 description 5
- 230000001174 ascending effect Effects 0.000 description 3
- 238000005286 illumination Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Abstract
The invention provides a method and a system for detecting dead pixels of an image sensor. The method comprises the following steps: using the image sensor to respectively collect monochrome images corresponding to plural monochrome panels of different colors; performing pixel analysis to each monochrome image, and recording the obtained dead pixel coordinates of each monochrome image in each monochrome dead pixel list; and comparing the monochrome dead pixel lists one another to comprehensively determine the dead pixel coordinates of the image sensor. The invention also provides the system for detecting dead pixels of an image sensor, which comprises a monochrome image obtaining module, a pixel analyzing module and a dead pixel determining module. The invention can detect dead pixels under multi-color environment, can detect dead pixels of different conditions more accurately, improves accuracy of dead pixel detection, analyzes the dead pixel list of each monochrome image comprehensively, divides different dead pixel grades, and can adapt to requirements of different detection quality with strong flexibility.
Description
Technical field
The present invention relates to technical field of image processing, particularly relate to the method and system that a kind of dead pixels of image sensor is surveyed.
Background technology
Current digital video and image product all need imageing sensor to gather original image.Due to technique and raw material restriction, imageing sensor often has bad point.So-called bad point, referring to not can personal comments light varience, presents all the time the pixel of same color (as white, black or color point), and it is insensitive to different input pictures, always exports fixing pixel value.Detection to bad point and processing can improve the quality of image greatly.
Under normal circumstances, dead pixels of image sensor survey method be: in the environment of complete darkness (for example, can obtain by covering lens cap the environment of complete darkness), utilize imageing sensor to obtain a monochromatic black image, then the black pixel value of choice criteria is as with reference to as numerical value, each pixel to this figure judges, if the absolute value of the difference of the pixel value of this pixel and reference pixel value is greater than threshold value, to putting the bad point that is judged as black image.In addition, can obtain the monochromatic white image under uniform illumination by imageing sensor, the white pixel value of choice criteria is as with reference to as numerical value, the pixel value of each pixel and reference pixel value are compared, the bad point that obtains white image, the detection method of its detection method and black image is similar.
The general detection method that adopts black image and white image to combine, can detect black bad point and white bad point.But for the pixel between black and white, for example grey bad point, the difference of its pixel value and black reference pixel value is less, and the difference of white reference pixel value is also less, therefore in testing process, may be judged as normal point, occurs undetected.
In a word, need at present the problem that those skilled in the art solve to be exactly: a kind of method that provides dead pixels of image sensor to survey, to avoid occurring the undetected situation of bad point.
Summary of the invention
Technical problem to be solved by this invention is to provide a kind of method that dead pixels of image sensor is surveyed, to solve the undetected situation of bad point.
In order to address the above problem, the invention discloses a kind of method that dead pixels of image sensor is surveyed, it is characterized in that, comprising:
Utilize imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
Respectively each monochrome image is carried out to pixel analysis, by the bad point coordinate record of each monochrome image obtaining in each monochromatic bad point list;
By relatively more described each monochromatic bad point list, comprehensively determine the bad point coordinate of described imageing sensor.
Further, described pixel analysis comprises:
Each pixel component value to monochrome image sorts respectively, obtains the component value sequence of each pixel component;
Determine benchmark pixel component value and compared pixels component value for each component value sequence;
Calculate benchmark pixel component value in each component value sequence and the difference of compared pixels component value, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
Contrast the suspicious bad point of each pixel component, will there is the suspicious bad point of same coordinate as the bad point of monochrome image.
Preferably, the definite of described benchmark pixel component value comprises:
The same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value; Or, the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value.
Preferably, the definite of described compared pixels component value comprises:
Pixel component value except benchmark pixel component value in described component value sequence is defined as to compared pixels component value;
Or, according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value.
Preferably, the monochrome panel of described different colours comprises: black panel, white panel, red panel, green panel and blue panel.
Preferably, described monochrome image comprises: RGB pattern, YUV pattern, Lab pattern.
The system that the present invention also provides a kind of dead pixels of image sensor to survey, is characterized in that, comprising:
Monochrome image acquisition module, for utilizing imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
Pixel analysis module, for respectively each monochrome image being carried out to pixel analysis, by the bad point coordinate record of each monochrome image obtaining in each monochromatic bad point list;
Bad point determination module, for by relatively more described each monochromatic bad point list, comprehensively determines the bad point coordinate of described imageing sensor.
Further, described pixel analysis module comprises:
Component value sequencing unit, sorts respectively for each pixel component value to monochrome image, obtains the component value sequence of each pixel component;
Component value determining unit, for determining benchmark pixel component value and compared pixels component value for each component value sequence;
Suspicious bad point acquiring unit, for calculating the difference of benchmark pixel component value and compared pixels component value of each component value sequence, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
Monochromatic bad point acquiring unit, for contrasting the suspicious bad point of each pixel component, will have the suspicious bad point of same coordinate as the bad point of monochrome image.
Preferably, described component value determining unit comprises:
Benchmark pixel component value is determined subelement, for the same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value; Or, the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value.
Compared pixels component value is determined subelement, for by described component value sequence, the pixel component value except benchmark pixel component value is defined as compared pixels component value; Or, according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value.
Preferably, the monochrome panel of described each different colours comprises: black panel, white panel, red panel, green panel and blue panel
Compared with prior art, the present invention has the following advantages:
The present invention makes imageing sensor pass through different monochrome panels, and the monochrome image of the different colours collecting is carried out to pixel analysis processing, obtains the bad point list of monochrome image, and the bad point of last comprehensive each monochrome image is determined the bad point coordinate of imageing sensor.The present invention utilizes multiple monochrome images to detect respectively, can detect the multiple bad points such as stain, white point and color point, can detect more exactly the bad point of different situations, avoid in black and white image detection, the generation of the undetected phenomenon of bad point, has improved the accuracy of bad point detection.
Simultaneously, the present invention can comprehensively analyze the bad point list of each monochrome image, the number of times occurring in the bad point list of each monochrome image according to the coordinate of pixel, divide different bad point ranks, for example, the coordinate all occurring in the bad point list of each monochrome image, its corresponding pixel one is decided to be bad point; At few pixel corresponding to coordinate of the bad point list occurrence number of each monochrome image, can, according to user's different needs, in detection, be ignored, be considered as normal pixel.The present invention can adapt to the different needs that detect quality, and flexibility is strong.
Brief description of the drawings
Fig. 1 is the flow chart of the embodiment of the method for a kind of dead pixels of image sensor survey of the present invention;
Fig. 2 is the flow chart of the method preferred embodiment of a kind of dead pixels of image sensor survey of the present invention;
Fig. 3 is the structure chart of the system embodiment of a kind of dead pixels of image sensor survey of the present invention.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, below in conjunction with the drawings and specific embodiments, the present invention is further detailed explanation.
Because the bad point of imageing sensor is insensitive to the image of input, always can export more fixing pixel value, in the detection of monochrome image, may there is situation about being more or less the same with reference pixel value in the pixel value of bad point, and occur undetected.The present invention detects the method for bad point to monochrome image according to prior art, utilize transducer to gather multiple monochrome images to the monochrome panel of multiple different colours, detect by the bad point to each monochrome image respectively, bad point under comprehensive comparative analysis different colours environment, thereby draw the bad point of imageing sensor, can detect more exactly the bad point of imageing sensor.
With reference to Fig. 1, show the flow chart of the embodiment of the method for a kind of dead pixels of image sensor survey of the present invention, comprising:
Step 101, utilizes imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
For imageing sensor to be detected, the monochrome panel of the multiple different colours of its alignment lens is taken pictures, gather corresponding monochrome image.The monochrome panel that adopts different colours, can obtain the photosensitive property of imageing sensor to different colours, thereby obtain the bad point of different colours environment hypograph transducer.The monochrome panel of described different colours can comprise black panel, white panel and other color panel, and choosing of panel color can be chosen according to different examination criterias.If the detection quality requirement of bad point is higher, can choose the panel of a greater number different colours; If the detection quality of bad point otherwise lower, can choose the panel of lesser amt different colours.
Preferably, the monochrome image collecting can be the image of RGB pattern, YUV pattern, Lab pattern.RGB pattern is that R represents red R ed by variation and their stacks each other red, green, blue three Color Channels are obtained to color miscellaneous, and G represents green Green, and B represents blue Blue.Lab pattern is by illumination L with about a of color,, tri-key elements compositions of b.L represents illumination (Luminosity), is equivalent to brightness, and a represents that, from redness to green scope, b represents from yellow to blue scope.In YUV colour model, Y represents brightness, and U and V represent aberration, forms two colored components.
Step 102, sorts respectively to each pixel component value of each monochrome image, obtains the component value sequence of each pixel component;
Carry out different sequences according to the pattern difference of collected monochrome image, for each monochrome image, the image of RGB pattern if, sorts to R component value, G component value and B component value respectively; The image of YUV pattern if, sorts to Y component value, U component value and V component value respectively; The image of Lab pattern if, sorts to L component value, a component value and b component value respectively.Described sort method has multiple, as bubble sort method, selection sort, insertion sort etc.This step obtains the component value sequence of each pixel component of its correspondence for each monochrome image.
Step 103, determines benchmark pixel component value and compared pixels component value for each component value sequence;
Each component value sequence for each monochrome is determined the benchmark pixel component value corresponding with pixel component and compared pixels component value.Determining of described benchmark pixel component value can be to be positioned at the middle same pixel component value of described component value sequence.Because the pixel of the transducer overwhelming majority is normal pixel point, the pixel component value of normal pixel point is positioned in the middle of component value sequence and accounts for the overwhelming majority, therefore the same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value.Can also be that the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value.Because the pixel component value of bad point is greater than or less than the pixel component value of normal pixel point, therefore the mean value of each component value order also can be used as the judgment standard of bad point.
Determining of described compared pixels component value can be selected the pixel component value except benchmark pixel component value in described component value sequence.Can also according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value, wherein, choose different pre-set criteria according to different detection quality, if the detection quality requirement of bad point is higher, choose in component value sequence before sequence with sequence after most component value; If the detection quality requirement of bad point is lower, choose before sequence in component value sequence with sequence after the component value of fraction.
Step 104, calculates benchmark pixel component value in each component value sequence and the difference of compared pixels component value, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
For each pixel component of each monochrome image to there being different preset bad point threshold values, the setting of described preset bad point threshold value is chosen different values according to different detection quality, for example, for the R component value sequence of black image, if set higher detection quality, can choose preset bad point threshold value is 5, and benchmark pixel component value and the absolute value of the difference of compared pixels component value are greater than 5 o'clock corresponding pixels and are judged as suspicious bad point; If set lower detection quality, can choose preset bad point threshold value is 10, and benchmark pixel component value and the absolute value of the difference of compared pixels component value are greater than 10 o'clock corresponding pixels and are judged as suspicious bad point.
Step 105, contrasts the suspicious bad point of each pixel component, will have the suspicious bad point of same coordinate as the bad point of monochrome image, and by its coordinate record in each monochromatic bad point list;
There is the suspicious bad point of each pixel component of same coordinate as the bad point of monochrome image, in corresponding color environment, each pixel component value of this point does not have normal standard, and it is undesired that this point shows in corresponding monochrome image, is the bad point of monochrome image.Record the coordinate of this pixel, corresponding different color is recorded in different monochromatic bad point lists.
Step 106, by relatively more described each monochromatic bad point list, comprehensively determines the bad point coordinate of described imageing sensor.
The bad point coordinate of described imageing sensor is determined according to comprehensive analysis of each monochromatic bad point list.The number of times occurring in each monochromatic bad point list according to monochromatic bad point coordinate, divides different bad point grades.The coordinate all occurring in the bad point list of each monochrome image, its corresponding pixel one is decided to be bad point, can be judged as senior bad point; There is pixel corresponding to coordinate once in the bad point list of each monochrome image, can be judged as one-level bad point; There is the pixel corresponding to coordinate of twice, can be judged as secondary bad point, and by that analogy.According to the needs of user's different application condition, choose the bad point of appropriate level as the bad point of imageing sensor, remaining monochromatic bad point can be ignored in detection, is considered as normal pixel.
The method of surveying according to above-mentioned a kind of dead pixels of image sensor for the monochrome image of each different colours collecting, can be analyzed to each monochrome image parallel processing simultaneously; Also can analyze respectively successively serial process to the image of each color.Preferably, for improving the efficiency of bad point detection, each monochrome image is carried out to analyzing and processing simultaneously.
The embodiment of the present invention utilizes multiple monochrome images to detect respectively, can detect the bad point under multiple color environment, compared with the bad point detection of solid color, can detect more exactly bad point, avoid the generation of the undetected phenomenon of bad point in black and white image detection, improved the accuracy of bad point detection.In addition, the present invention comprehensively analyzes the bad point list of each monochrome image, and the number of times occurring in the bad point list of each monochrome image according to the coordinate of pixel, divides different bad point ranks, can adapt to the different needs that detect quality, and flexibility is strong.
With reference to Fig. 2, show the flow chart of the method preferred embodiment of a kind of dead pixels of image sensor survey of the present invention, comprising:
Step 201, initialisation image transducer;
Step 202, utilizes the imageing sensor corresponding monochrome image of monochrome panel collection to black, white, redness, green, blueness successively;
The camera lens of imageing sensor is aimed at successively the monochrome panel of black, white, redness, green, blueness and is taken pictures, and gathers the monochrome image of corresponding black, white, redness, green, blueness.The pattern of described testing image is RGB pattern.
Step 203, R component value, G component value and B component value to each monochrome image carry out respectively bubble sort, obtain the component value sequence of each pixel component;
R component value for black image carries out bubble sort, obtains the R component value sequence of the arrangement ascending according to R component value, in like manner also obtains according to G component value sequence and the B component value sequence of G component value and the ascending arrangement of B component value.
Meanwhile, for white, red, green, blue image, obtain R component value sequence, G component value sequence and the B component value sequence of corresponding monochrome image.
Step 204, the same pixel component value in the middle of being positioned at described component value sequence is as benchmark pixel component value, by the pixel component value as a comparison of the pixel component value except benchmark pixel component value in described component value sequence;
For example, the R component value sequence of black image is:
[236，240，242，245，245，245，245，245，245，245，245，246，252，255]
G component value sequence is:
[237，243，246，247，247，247，247，247，247，247，247，249，252，255]
B component value sequence is:
[236，240，242，245，245，245，245，245，245，245，245；246，252，255]
Choose the 245 benchmark R component values as black image, 236,240,242,246,252, the 255 comparison G component values as black image; Choose the 247 benchmark G component values as black image, 237,243,246249,252, the 255 comparison component values as black image; Choose the 245 benchmark B component values as black image, 236,240,242,246,252, the 255 comparison B component values as black image.
In like manner, the definite and black image of the benchmark pixel component value of white, redness, green, blue image and compared pixels component value is definite similar.
Step 205, calculates benchmark pixel component value in each component value sequence and the difference of compared pixels component value, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
In the present embodiment.For the R component value sequence of black image, benchmark R component value is respectively 9,5,3,1,6,10 with the difference that compares R component value.Described preset bad point threshold value gets 4, judges that difference is that the point of 9,5,6,10 correspondences is suspect pixel point, is designated as suspicious bad point a, b, c, d herein.
For the G component value sequence of black image, benchmark G component value is respectively 10,4,1,2,5,8 with the difference that compares G component value.Described preset bad point threshold value gets 4, judges that difference is that the point of 10,5,8 correspondences is suspect pixel point, is designated as suspicious bad point e, f, g herein.
For the B component value sequence of black image, benchmark B component value is respectively 9,6,2,2,5,10 with the difference that compares B component value.Described preset bad point threshold value gets 4, judges that difference is that the point of 9,6,5,10 correspondences is suspect pixel point, is designated as suspicious bad point h, i, j, k herein.
Accordingly, the judgement of the suspicious bad point of each pixel component of white, redness, green, blue image is same as above.
Step 206, for each monochrome image, is recorded in the coordinate of the suspicious bad point of R, G, B component in corresponding R, G, the list of B bad point by the ascending of the absolute value of described difference;
According to the size of the absolute value of difference corresponding to pixel in step 205, the order of pressing b, c, a, d by the coordinate record of corresponding points in the list of R bad point, the order of pressing f, g, e by the coordinate record of corresponding points in the list of G bad point, by the order of j, i, h, k by the coordinate record of corresponding points in the list of G bad point.
Accordingly, white, red, green, blue image R, G, the list of B bad point record same as above.
Step 207, contrast R, G, the list of B bad point, will have the suspicious bad point of same coordinate as the bad point of monochrome image, and by its coordinate record in each monochromatic bad point list;
For black image, contrast R, G, the list of B bad point, in the present embodiment, if some a, e, h have identical coordinate, be designated as an A; Point b, g, k have identical coordinate, are designated as a B; Point c, g, i has identical coordinate, is designated as a C, the bad point that to put A, B, C be black image, and A, B, C are recorded in the list of black bad point.
Accordingly, choose the bad point of white, redness, green, blue image and be recorded in corresponding monochromatic bad point list.
Step 208, by relatively more described each monochromatic bad point list, comprehensively determines the bad point coordinate of described imageing sensor.
The relatively monochromatic bad point list of black, white, redness, green, blue image, according to the bad point coordinate of the standard imageing sensor detecting.In the present embodiment, the coordinate of postulated point A occurred five times in described five monochromatic bad point lists, and the coordinate of some B occurred four times in described five monochromatic bad point lists, and the coordinate of some C has only occurred once in the list of black bad point.If examination criteria is had relatively high expectations, be decided to be the coordinate occurring all in monochromatic bad point list, the bad point that judging point A, B, C are imageing sensor; If examination criteria is to occur more than three times coordinates in monochromatic bad point list, judging point A and the some B bad point that is imageing sensor; If it is lower that examination criteria requires, be decided to be the coordinate all occurring in monochromatic bad point list, the bad point that judging point A is imageing sensor.The judgement of other pixels is described above.
According to the embodiment of the present invention, the bad point list of each monochrome image is comprehensively analyzed, judge that the coordinate of pixel is at the number of times of the bad point list appearance of each monochrome image, thereby the bad point of judging imageing sensor, flexibility is strong, can meet the needs of different examination criterias.
With reference to Fig. 3, show the structure chart of the system embodiment of a kind of dead pixels of image sensor survey of the present invention, it is characterized in that, comprising:
Monochrome image acquisition module 301, for utilizing imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
Pixel analysis module 302, for respectively each monochrome image being carried out to pixel analysis, by the bad point coordinate record of each monochrome image obtaining in each monochromatic bad point list;
Bad point determination module 303, for by relatively more described each monochromatic bad point list, comprehensively determines the bad point coordinate of described imageing sensor.
Further, described pixel analysis module 302 comprises:
Component value sequencing unit 3021, sorts respectively for each pixel component value to monochrome image, obtains the component value sequence of each pixel component;
Component value determining unit 3022, for determining benchmark pixel component value and compared pixels component value for each component value sequence;
Suspicious bad point acquiring unit 3023, for calculating the difference of benchmark pixel component value and compared pixels component value of each component value sequence, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
Monochromatic bad point acquiring unit 3024, for contrasting the suspicious bad point of each pixel component, will have the suspicious bad point of same coordinate as the bad point of monochrome image.
Further, described component value determining unit 3022 comprises:
Benchmark pixel component value is determined subelement 30221, for the same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value; Or, the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value.
Compared pixels component value is determined subelement 30222, for by described component value sequence, the pixel component value except benchmark pixel component value is defined as compared pixels component value; Or, according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value.
Preferably, the monochrome panel of described each different colours comprises: black panel, white panel, red panel, green panel and blue panel.
Monochrome image acquisition module 301 is by the corresponding monochrome image of monochrome panel collection of above-mentioned each different colours.Pixel analysis module 302 is analyzed above-mentioned monochrome image, first each pixel component value of monochrome image is compared to judgement, draw the suspicious bad point of respective components value, then contrast the suspicious bad point of each pixel component, the bad point that obtains monochrome image, is then recorded in each monochromatic bad point list.Finally, relatively more described each the monochromatic bad point list of bad point determination module 303, according to the needs of difference detection quality, comprehensively judges, determines the bad point coordinate of imageing sensor.
Each embodiment in this specification all adopts the mode of going forward one by one to describe, and what each embodiment stressed is and the difference of other embodiment, between each embodiment identical similar part mutually referring to.For system embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, relevant part is referring to the part explanation of embodiment of the method.
The method and system of above a kind of dead pixels of image sensor provided by the present invention being surveyed, be described in detail, applied specific case herein principle of the present invention and execution mode are set forth, the explanation of above embodiment is just for helping to understand method of the present invention and core concept thereof; , for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention meanwhile.
Claims (8)
1. the method that dead pixels of image sensor is surveyed, is characterized in that, comprising:
Utilize imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
Respectively each monochrome image is carried out to pixel analysis, by the bad point coordinate record of each monochrome image obtaining in each monochromatic bad point list;
By relatively more described each monochromatic bad point list, comprehensively determine the bad point coordinate of described imageing sensor;
Wherein, described pixel analysis comprises:
Each pixel component value to monochrome image sorts respectively, obtains the component value sequence of each pixel component;
Determine benchmark pixel component value and compared pixels component value for each component value sequence;
Calculate benchmark pixel component value in each component value sequence and the difference of compared pixels component value, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
Contrast the suspicious bad point of each pixel component, will there is the suspicious bad point of same coordinate as the bad point of monochrome image.
2. method according to claim 1, is characterized in that, the definite of described benchmark pixel component value comprises:
The same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value; Or, the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value.
3. method according to claim 2, is characterized in that, the definite of described compared pixels component value comprises:
Pixel component value except benchmark pixel component value in described component value sequence is defined as to compared pixels component value;
Or, according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value.
4. method according to claim 1, is characterized in that, the monochrome panel of described different colours comprises: black panel, white panel, red panel, green panel and blue panel.
5. method according to claim 1, its feature exists, and described monochrome image comprises: RGB pattern, YUV pattern, Lab pattern.
6. the system that dead pixels of image sensor is surveyed, is characterized in that, comprising:
Monochrome image acquisition module, for utilizing imageing sensor to gather respectively corresponding monochrome image to the monochrome panel of multiple different colours;
Pixel analysis module, for respectively each monochrome image being carried out to pixel analysis, by the bad point coordinate record of each monochrome image obtaining in each monochromatic bad point list;
Bad point determination module, for by relatively more described each monochromatic bad point list, comprehensively determines the bad point coordinate of described imageing sensor;
Wherein, described pixel analysis module comprises:
Component value sequencing unit, sorts respectively for each pixel component value to monochrome image, obtains the component value sequence of each pixel component;
Component value determining unit, for determining benchmark pixel component value and compared pixels component value for each component value sequence;
Suspicious bad point acquiring unit, for calculating the difference of benchmark pixel component value and compared pixels component value of each component value sequence, if the absolute value of described difference is greater than preset bad point threshold value, corresponding pixel points is judged as suspicious bad point;
Monochromatic bad point acquiring unit, for contrasting the suspicious bad point of each pixel component, will have the suspicious bad point of same coordinate as the bad point of monochrome image.
7. system according to claim 6, is characterized in that, described component value determining unit comprises:
Benchmark pixel component value is determined subelement, for the same pixel component value being positioned in the middle of described component value sequence is defined as to benchmark pixel component value; Or, the pixel component value of described component value sequence is averaging, the average pixel value obtaining is defined as to benchmark pixel component value;
Compared pixels component value is determined subelement, for by described component value sequence, the pixel component value except benchmark pixel component value is defined as compared pixels component value; Or, according to pre-set criteria by before sequence in described component value sequence with sequence after component value be defined as compared pixels component value.
8. system according to claim 6, is characterized in that, the monochrome panel of described each different colours comprises: black panel, white panel, red panel, green panel and blue panel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910238438.2A CN101715050B (en) | 2009-11-20 | 2009-11-20 | Method and system for detecting dead pixels of image sensor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910238438.2A CN101715050B (en) | 2009-11-20 | 2009-11-20 | Method and system for detecting dead pixels of image sensor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101715050A CN101715050A (en) | 2010-05-26 |
| CN101715050B true CN101715050B (en) | 2014-07-09 |
Family
ID=42418303
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910238438.2A Active CN101715050B (en) | 2009-11-20 | 2009-11-20 | Method and system for detecting dead pixels of image sensor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101715050B (en) |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101895786B (en) * | 2010-07-15 | 2012-07-25 | 杭州海康威视软件有限公司 | Detection method and device for image sensor |
| CN104599995B (en) * | 2015-01-14 | 2017-04-05 | 上海华岭集成电路技术股份有限公司 | The offline method for positioning the continuous bad point of image sensor chip |
| CN105987807A (en) * | 2015-02-15 | 2016-10-05 | 宁波舜宇光电信息有限公司 | Defective point detection system and method for vertical cavity surface emitting laser array |
| CN104700424B (en) * | 2015-03-30 | 2017-07-11 | 山东省计量科学研究院 | Medical color fujinon electronic video endoscope dead pixel points of images detection means |
| CN104935837B (en) * | 2015-06-10 | 2018-02-06 | 无锡英斯特微电子有限公司 | For optical imagery pointing device and the image processing method of optical video equipment |
| CN105430385B (en) * | 2015-12-14 | 2017-06-30 | 上海富瀚微电子股份有限公司 | The method and device that a kind of dead pixels of image sensor is surveyed and corrected |
| CN105704406B (en) * | 2016-02-01 | 2018-12-18 | 上海集成电路研发中心有限公司 | A kind of method of bad point removal in image procossing |
| CN106878636B (en) * | 2017-02-13 | 2019-11-29 | 建荣半导体（深圳）有限公司 | Defect pixel point detecting method, device and picture processing chip |
| CN107016669B (en) * | 2017-03-27 | 2019-08-23 | 福州瑞芯微电子股份有限公司 | A kind of dead pixel points of images detection method and device |
| CN109660788B (en) * | 2017-10-10 | 2020-12-22 | 凌云光技术股份有限公司 | Bad cluster detection method and device |
| CN107635087A (en) * | 2017-11-08 | 2018-01-26 | 深圳地平线机器人科技有限公司 | Solid state image sensor and image-sensing method |
| CN108259892A (en) * | 2018-03-21 | 2018-07-06 | 深圳怡化电脑股份有限公司 | Detection method, system and the device of imaging sensor |
| WO2019196009A1 (en) * | 2018-04-10 | 2019-10-17 | 深圳市大疆创新科技有限公司 | Method for detecting defective pixel of image sensor, and photography apparatus, unmanned aerial vehicle and storage medium |
| CN109959498A (en) * | 2019-03-29 | 2019-07-02 | 珠海达明科技有限公司 | A kind of array light-emitting component quality detecting system and its application method |
| CN110275794B (en) * | 2019-06-27 | 2022-07-22 | 努比亚技术有限公司 | Screen display method of terminal, and computer-readable storage medium |
| CN110708537B (en) * | 2019-09-30 | 2021-05-28 | 芯盟科技有限公司 | Image sensor performance testing method and device and storage medium |
| CN113014908A (en) * | 2019-12-19 | 2021-06-22 | 西安诺瓦星云科技股份有限公司 | Image detection method, device and system and computer readable storage medium |
| CN110830795B (en) * | 2020-01-09 | 2020-05-19 | 武汉精立电子技术有限公司 | Image sensor dead pixel detection method and device and electronic equipment |
| CN113630594B (en) * | 2021-08-04 | 2022-05-17 | 蚌埠市高远光电有限公司 | Bad point detection system of display panel |
| CN114727102A (en) * | 2022-06-10 | 2022-07-08 | 江苏游隼微电子有限公司 | Driving-assistant camera dead pixel correction optimization method |
| CN115526836A (en) * | 2022-08-30 | 2022-12-27 | 中国软件评测中心(工业和信息化部软件与集成电路促进中心) | Image sensor dead pixel detection method, system, storage medium and electronic device |
| CN116506719B (en) * | 2023-06-21 | 2023-09-15 | 深圳华强电子网集团股份有限公司 | Transmission management method based on photodiode CMOS image sensor |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101119502A (en) * | 2007-08-22 | 2008-02-06 | 北京中星微电子有限公司 | Image processing device and image processing method |
| CN101150658A (en) * | 2007-08-22 | 2008-03-26 | 北京中星微电子有限公司 | A bad point self-adapted grid noise elimination device and method |
| CN101242542A (en) * | 2008-02-26 | 2008-08-13 | 北京中星微电子有限公司 | An image detection method and device |
| CN101304485A (en) * | 2008-05-26 | 2008-11-12 | 北京中星微电子有限公司 | Method and apparatus for compensating bad point |
-
2009
- 2009-11-20 CN CN200910238438.2A patent/CN101715050B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101119502A (en) * | 2007-08-22 | 2008-02-06 | 北京中星微电子有限公司 | Image processing device and image processing method |
| CN101150658A (en) * | 2007-08-22 | 2008-03-26 | 北京中星微电子有限公司 | A bad point self-adapted grid noise elimination device and method |
| CN101242542A (en) * | 2008-02-26 | 2008-08-13 | 北京中星微电子有限公司 | An image detection method and device |
| CN101304485A (en) * | 2008-05-26 | 2008-11-12 | 北京中星微电子有限公司 | Method and apparatus for compensating bad point |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101715050A (en) | 2010-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101715050B (en) | Method and system for detecting dead pixels of image sensor | |
| CN1960439B (en) | Methods and devices for image signal processing | |
| CN101080022B (en) | Imaging device | |
| CN100341313C (en) | Method of determining color composition of an image | |
| CN109471276B (en) | Method and device for detecting color cast defect of liquid crystal display | |
| CN101896865A (en) | Device for evaluating the surface of a tyre | |
| CN105303525B (en) | Haze image discrimination apparatus and discriminating conduct | |
| US8379975B2 (en) | Image processing device and method, and program recording medium | |
| CN106937064A (en) | The detection method of digital picture bad point, bearing calibration and device | |
| US8929651B2 (en) | System and method for the repair of anomalies in images | |
| CN107767390A (en) | The shadow detection method and its system of monitor video image, shadow removal method | |
| CN103186905A (en) | Color detector for vehicle | |
| CN105509657A (en) | High-precision measurement method of scratch area and grayscale parameter of IC card chip | |
| CN101939995A (en) | Image processing device and image processing program | |
| KR101608843B1 (en) | System and Method for Automatically Detecting a Mura Defect using Advanced Weber's Law | |
| JP2010008159A (en) | Visual inspection processing method | |
| US7327384B2 (en) | Gamut false color display | |
| CN110146509B (en) | Battery detection method and battery detection equipment | |
| US10083516B2 (en) | Method for segmenting a color image and digital microscope | |
| CN101178422B (en) | Character type illuminating device detecting method | |
| TWI313848B (en) | A fire image detection method | |
| CN105744260B (en) | A kind of bar paten detection method and device | |
| KR102655139B1 (en) | Camera module false feature classification method and system according to image segmentation inspection | |
| JP2755807B2 (en) | Burner clogging identification method | |
| Minagawa et al. | A color chart detection method for automatic color correction |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20180408 Address after: 100191 Xueyuan Road, Haidian District, Haidian District, Beijing, No. 607, No. six Patentee after: Beijing Vimicro AI Chip Technology Co Ltd Address before: 100083 Haidian District, Xueyuan Road, No. 35, the world building, the second floor of the building on the ground floor, No. 16 Patentee before: Beijing Vimicro Corporation |
|
| TR01 | Transfer of patent right |