CN115002364A - Method and device for detecting dead pixel - Google Patents
Method and device for detecting dead pixel Download PDFInfo
- Publication number
- CN115002364A CN115002364A CN202210529149.3A CN202210529149A CN115002364A CN 115002364 A CN115002364 A CN 115002364A CN 202210529149 A CN202210529149 A CN 202210529149A CN 115002364 A CN115002364 A CN 115002364A
- Authority
- CN
- China
- Prior art keywords
- pixel
- value
- dead
- pixels
- current
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
- H04N17/002—Diagnosis, testing or measuring for television systems or their details for television cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for detecting a dead pixel, which comprises the following steps: detecting suspected dead pixels in the previous frame of image; judging whether the suspected dead pixel is a real dead pixel or not by utilizing the motion background according to the current frame image; writing the bad point mark value into the storage of the known bad point, and processing the bad point as a static bad point in the next frame image; the invention also discloses a device for detecting the dead pixel, which comprises a detection unit, a calculation unit and a judgment unit. The invention finds out the more abrupt image strong boundary in the previous frame, marks the image strong boundary as a suspected dead pixel area, and then utilizes the characteristics of larger difference value of moving images of the previous frame and the next frame and smaller change of the dead pixel to identify the real dead pixel.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a device for detecting dead pixels.
Background
Some bad spots inevitably exist in the manufactured image sensor due to the difference between the semiconductor manufacturing process and the raw material. The dead pixel refers to a pixel point which can not always present the same brightness along with the change of the light sensitivity; the presence of the defective dots is one of the causes of the degradation of the image quality, and therefore, the detection and processing of the defective dots are very necessary to improve the image quality.
At present, when a dead pixel of an image sensor is detected in the prior art, a fixed pixel threshold is usually set, and for images acquired under different scenes, whether a pixel point is a dead pixel is judged by comparing a certain statistic with the pixel threshold.
Disclosure of Invention
The invention aims to solve the problem that the accuracy of a detection result is low due to the fact that missing detection and misjudgment of dead pixels are easily caused in the prior art, and provides a dead pixel detection method and a dead pixel detection device.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method of detecting a dead pixel, the method comprising:
s1, detecting suspected dead pixels in a previous frame of image;
s2, calculating a threshold value and a pixel difference value between two frames at two sides of a suspected dead pixel according to the current frame image;
and S3, judging whether any pixel value is larger than a threshold value, if one pixel difference value is larger than a dead pixel threshold value, judging that the pixel is a real dead pixel, writing a dead pixel mark value into the storage of the known dead pixel, and processing the dead pixel mark value as a static dead pixel in the RAW image of the next frame, otherwise, executing the step S2 until all suspected dead pixels are detected.
Preferably, the step S1 specifically includes:
s11, acquiring a pixel block with the size of N multiplied by N taking a current pixel point as a center in a previous frame of image;
s12, calculating an average value, a minimum value, a maximum value and an average value of difference values in the pixel blocks, and calculating a threshold value for judging a boundary difference value of the suspected dead pixel;
s13, calculating the minimum difference value of all pixel points in the pixel block in different preset directions;
s14, judging whether the minimum difference value in a certain preset direction is larger than a threshold value or not, if the minimum difference value in the certain preset direction is larger than the threshold value, indicating that the boundary is a possible dead pixel boundary, counting the number of the boundaries, initializing the mark value of the suspected dead pixel as a, and if the number of the boundaries is zero, indicating that the difference value between the current pixel and the surrounding pixels is small and the pixel is a normal pixel;
if the nearest pixel in a certain direction is a known dead pixel and the current pixel value is closer to the pixel value of the known dead pixel, judging that the current pixel is a suspected dead pixel, and setting a mark value as b;
if the pixel value of the current pixel point is close to the maximum value or the minimum value in the pixel block and the number of the dead pixel boundaries is greater than a preset value, setting a marking value as b;
and storing the mark value of the same line in a memory, storing the pixels of the current line in a line cache, and performing interframe judgment on the pixels in the next frame.
Preferably, the step S2 specifically includes:
s21, obtaining pixels of the same line in the current frame image corresponding to a line of pixels of a previous frame for judging suspected dead pixels, defining a current pixel point as x, extending M pixels to the left side by taking the current pixel point x as a center, extending N pixels to the right side to form an inter-frame block, and forming a left boundary xl and a right boundary xr of the inter-frame in the current line;
s22, calculating the minimum value, the maximum value, the average difference value and the pixel average value in the inter-frame block, and calculating a dead pixel threshold value for judging whether the inter-frame block is static or moving;
and S23, judging whether the suspected dead pixel is a real dead pixel.
Preferably, the preset directions in step S13 include four preset directions, i.e., left, upper, right, and lower.
An apparatus for detecting a dead pixel, the apparatus comprising:
the detection unit is used for detecting suspected dead pixels in the previous frame of image;
the calculation unit is used for calculating a threshold value and a pixel difference value between two frames at two sides of the suspected dead pixel according to the current frame image;
a judging unit for judging whether any one of the pixel values is larger than a threshold value.
Preferably, the detection unit includes:
the first acquisition module is used for acquiring a pixel block with the size of N multiplied by N taking a current pixel point as a center in a previous frame image;
the first calculation module is used for calculating an average value, a minimum value, a maximum value and an average value of difference values in the pixel blocks and calculating a threshold value for judging a boundary difference value of the suspected dead pixel;
the second calculation module is used for calculating the minimum difference values of all pixel points in the pixel block in different preset directions;
the first judgment module is used for judging whether the minimum difference value in a certain preset direction is larger than a threshold value or not.
Preferably, the calculation unit includes:
a second obtaining module, configured to obtain pixels in the same row in the current frame image corresponding to a row of pixels in the previous frame for determining a suspected dead pixel, define a current pixel point as x, extend M pixels to the left side and extend N pixels to the right side with the current pixel point x as a center, so as to form an inter-frame block, and form a left boundary xl and a right boundary xr in the current row between frames;
a third calculation module for calculating a minimum value, a maximum value, an average difference value, and a pixel average value within an inter block;
and the second judging module is used for judging whether the suspected dead pixel is a real dead pixel or not.
Compared with the prior art, the invention provides a method and a device for detecting dead pixels, which have the following beneficial effects:
1. the invention finds out the more abrupt image strong boundary in the previous frame and marks the boundary as a suspected dead pixel area, and then identifies the real dead pixel in the next frame by utilizing the characteristics of larger difference value of the moving images of the previous frame and the next frame and smaller change of the dead pixel.
2. The method and the device can quickly detect the large bad points in the image by utilizing the characteristic that the bad points in the moving image background cannot be changed or are changed less, take the requirements of saving hardware resources into consideration, enhance the practicability and have higher market value and application prospect.
Drawings
FIG. 1 is a flow chart of a method for detecting a dead pixel according to the present invention;
FIG. 2 is a sub-flowchart of step S1 of a method for detecting a dead pixel according to the present invention;
FIG. 3 is a sub-flowchart of step S2 of a method for detecting a dead pixel according to the present invention;
FIG. 4 is a schematic diagram of a pixel block 5x5 according to a method for detecting a dead pixel of the present invention;
fig. 5 is a block diagram of a device for detecting a dead pixel according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In the description of the present invention, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention.
Referring to fig. 1 to 4, a method for detecting a dead pixel, where an image in this embodiment is an image in RAW format, includes:
s1, detecting suspected dead pixels in a previous frame RAW image, and specifically comprising the following steps:
s11, in the RAW image of the previous frame, taking the current pixel RawImag (y, x) as the center, obtaining a pixel block BlockBuf with size N × N on the image, in this embodiment, taking a pixel block with size 5 × 5 (shown in fig. 1), where the current pixel is BlockBuf (0,0), and marking as currpix.
The known dead pixels are noted as BlockDps for BlockBuf, the known dead pixels for each sensor can be externally input and stored, and BlockBuf and BlockDps are updated one by one with the center pixel.
S12, calculating an average value blk _ avg, a minimum value blk _ min, a maximum value blk _ max and an average value blk _ dif _ avg of the difference values in the pixel block Block Buf, wherein the calculation formula is as follows:
calculating a threshold thd _ sdp _ dif _ min for judging the boundary difference value of the suspected dead pixel, wherein the calculation formula is as follows: thd _ sdp _ dif _ min ═ blk _ avg/2+ blk _ dif _ avg × 2.
S13, calculating the minimum difference value dead _ pdiff [ i ] of all pixel points in the pixel block BlockBuf in four preset directions, namely left, upper, right, lower and the like, wherein the calculation formula is as follows:
dead_pdiff[0]＝min(|curPix-BlockBuf(0,-1)|,|curPix-BlockBuf(0,-2)|)；
dead_pdiff[1]＝min(|curPix-BlockBuf(-1,0)|,|curPix-BlockBuf(-2,0)|)；
dead_pdiff[2]＝min(|curPix-BlockBuf(0,1)|,|curPix-BlockBuf(0,2)|)；
dead_pdiff[3]＝min(|curPix-BlockBuf(1,0)|,|curPix-BlockBuf(2,0)|)；
if the minimum difference value dead _ pdiff [ i ] in a certain preset direction is greater than the threshold thd _ sdp _ dif _ min, which represents a possible dead point boundary flag _ pdiff [ i ], then the number of statistical boundaries sum _ flag _ pdiff is shown as follows:
flag_pdiff[i]＝(dead_pdiff[i]>thd_sdp_dif_min1:0)
initializing a mark value of the suspected dead pixel to be 0, namely, setting flag _ is _ sdp to be 0;
s14, if the sum _ flag _ pdiff is equal to 0, the difference value between the current pixel and the surrounding pixels is small, and the current pixel is a normal pixel;
if the nearest pixel in a certain direction is a known dead pixel and the current pixel value is closer to the pixel value of the known dead pixel, judging that the current pixel is a suspected dead pixel, and setting a flag _ is _ sdp to be 1;
if the pixel value of the current pixel point is close to the maximum value or the minimum value in the pixel block BlockBuf, and sum _ flag _ pdiff is more than or equal to 2, setting a flag _ is _ sdp to be 1;
storing the flag _ is _ sdp of the same line into a memory, recording as susdp (x) flag _ is _ sdp, storing the pixels of the current line into a line buffer LineBuf, and then performing inter-frame judgment on the pixels during the next frame;
s2, in the RAW image of the current frame, calculating a threshold value for judging whether the image is static or moving and a pixel difference value between two frames at two sides of a suspected dead pixel, and specifically comprising the following steps:
s21, obtaining pixels LineImg of the same line corresponding to a line of pixels which are judged to be suspected of being dead pixels in the previous frame from the current frame image, and defining the current pixel point as LineImg (x).
S22, with a current pixel point LineImg (x) as a center, extending M pixels to the left side and N pixels to the right side, in the embodiment, extending 3 pixels to the left side and 5 pixels to the right side to form an inter-frame block LineBuf with 8 pixels in total, and the left boundary xl and the right boundary xr of the inter-frame block LineBuf in a current line, then calculating the minimum value pix _ min, the maximum value pix _ max, the average difference value BlockDiffCur and the pixel average value pixsAvgCur in the inter-frame block LineBuf, and calculating a dead pixel threshold value for judging whether the image is static or moving;
s23, judging whether the suspected dead pixel is a real dead pixel or not, and specifically comprising the following steps:
s231, if the difference between the maximum value pix _ max and the minimum value pix _ min in the inter-frame block LineBuf is smaller than a set threshold, the suspected point area is considered to have no motion basically, and the step S232 needs to be executed;
s232. if | LineImg (xl) -LineImg (x1-2) | and | LineBuf (xl) -LineBuf (x1-2) | are both greater than a set threshold, then it indicates that there is a boundary of an object at xl, set flag _ edge _ lft ═ 1, otherwise flag _ edge _ lft ═ 0; similarly, if | LineImg (xr) -LineImg (xr +2) | and | LineBuf (xr) -LineBuf (xr +2) | are both greater than a set threshold, it indicates that there is an object boundary at xr, and set flag _ edge _ rgt to 1, otherwise, flag _ edge _ rgt to 0, and then step S233 needs to be executed;
s233, respectively calculating pixel difference values between two frames at two sides of the interval:
and S3, if one of blk _ dif _ lft and blk _ dif _ rgt is larger than a set threshold, the motion is obvious between the images of the two frames, the pixel in the current interval can be judged to be a real dead pixel, a dead pixel mark value is written into a storage of the known dead pixel and is treated as a static dead pixel in the RAW image of the next frame, and otherwise, the step S2 is executed until all suspected dead pixels are detected.
Referring to fig. 5, an apparatus for detecting a dead pixel, the apparatus comprising:
a detection unit 100, wherein the detection unit 100 is used for detecting a suspected dead pixel in a previous frame image, and the detection unit 100 comprises:
the first obtaining module 101, the first obtaining module 101 is configured to obtain a pixel block of an N × N size centered on a current pixel point in a previous frame of image;
the first calculating module 102, the first calculating module 102 is configured to calculate an average value, a minimum value, a maximum value, and an average value of difference values in the pixel blocks, and calculate a threshold value for determining a boundary difference value of the suspected dead pixel;
the second calculating module 103, the second calculating module 103 is configured to calculate minimum differences of all pixel points in the pixel block in different preset directions;
a first determining module 104, where the first determining module 104 is configured to determine whether a minimum difference value in a certain preset direction is greater than a threshold.
A calculating unit 200, wherein the calculating unit 200 is configured to calculate a threshold and a pixel difference between two frames at two sides of the suspected dead pixel according to the current frame image, and the calculating unit 200 includes:
a second obtaining module 201, where the second obtaining module 201 is configured to obtain pixels in a same row in the current frame image corresponding to a row of pixels in the previous frame for determining a suspected dead pixel, define a current pixel point as x, extend M pixels to the left side and N pixels to the right side with the current pixel point x as a center, and form an inter-frame block, and a left boundary xl and a right boundary xr in the current row between frames;
a third calculating module 202, wherein the third calculating module 202 is configured to calculate a minimum value, a maximum value, an average difference value and a pixel average value in the inter block;
the second judging module 203, the second judging module 203 is configured to judge whether the suspected dead pixel is a real dead pixel.
A judging unit 300, the judging unit 300 is used for judging whether any pixel value is larger than the threshold value.
In the embodiment, a more abrupt image strong boundary is found in the previous frame and is marked as a suspected dead pixel area, and then the real dead pixel is identified in the next frame by utilizing the characteristics that the moving image difference of the previous frame and the next frame is large, but the dead pixel change is small.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (7)
1. A method for detecting a dead pixel, the method comprising:
s1, detecting suspected dead pixels in a previous frame of image;
s2, calculating a threshold value and a pixel difference value between two frames at two sides of a suspected dead pixel according to the current frame image;
s3, judging whether any pixel value is larger than a threshold value, if any pixel difference value is larger than the threshold value, determining that the pixel is a real dead pixel, writing a dead pixel mark value into the storage of the known dead pixel, and processing the dead pixel as a static dead pixel in the next frame of RAW image, otherwise, executing the step S2 until all the suspected dead pixels are detected.
2. The method according to claim 1, wherein the step S1 specifically includes:
s11, acquiring a pixel block with the size of N multiplied by N taking a current pixel point as a center in a previous frame of image;
s12, calculating an average value, a minimum value, a maximum value and an average value of difference values in the pixel blocks, and calculating a threshold value for judging a boundary difference value of a suspected dead pixel;
s13, calculating the minimum difference value of all pixel points in the pixel block in different preset directions;
s14, judging whether the minimum difference value in a certain preset direction is larger than a threshold value or not, if the minimum difference value in the certain preset direction is larger than the threshold value, indicating that the boundary is a possible dead pixel boundary, counting the number of the boundaries, initializing the mark value of the suspected dead pixel as a, and if the number of the boundaries is zero, indicating that the difference value between the current pixel and the surrounding pixels is small and the pixel is a normal pixel;
if the nearest pixel in a certain direction is a known dead pixel and the current pixel value is closer to the pixel value of the known dead pixel, judging that the current pixel is a suspected dead pixel, and setting a mark value as b;
if the pixel value of the current pixel point is close to the maximum value or the minimum value in the pixel block and the number of the dead pixel boundaries is greater than a preset value, setting a marking value as b;
and storing the mark value of the same line in a memory, storing the pixels of the current line in a line buffer memory, and performing interframe judgment on the pixels in the next frame.
3. The method for detecting a dead pixel according to claim 1, wherein the step S2 specifically includes:
s21, obtaining pixels of the same line in the current frame image corresponding to a line of pixels of a previous frame for judging suspected dead pixels, defining a current pixel point as x, extending M pixels to the left side by taking the current pixel point x as a center, extending N pixels to the right side to form an inter-frame block, and forming a left boundary xl and a right boundary xr of the inter-frame in the current line;
s22, calculating the minimum value, the maximum value, the average difference value and the pixel average value in the inter-frame block, and calculating a dead pixel threshold value for judging whether the inter-frame block is static or moving;
and S23, judging whether the suspected dead pixel is a real dead pixel.
4. The method of claim 2, wherein the predetermined directions in step S13 include four predetermined directions, i.e., left, up, right, and down.
5. A device for detecting a defective pixel, the device comprising:
a detection unit (100), wherein the detection unit (100) is used for detecting suspected dead pixels in the previous frame image;
the calculating unit (200) is used for calculating a threshold value and a pixel difference value between two frames at two sides of the suspected dead pixel according to the current frame image;
a judging unit (300), wherein the judging unit (300) is used for judging whether any pixel value is larger than a threshold value.
6. A bad pixel detection apparatus according to claim 5, wherein said detection unit (100) comprises:
the first acquisition module (101) is used for acquiring a pixel block with the size of N multiplied by N taking a current pixel point as a center in a previous frame image;
a first calculating module (102), wherein the first calculating module (102) is used for calculating an average value, a minimum value, a maximum value and an average value of difference values in a pixel block and calculating a threshold value for judging a suspected dead pixel boundary difference value;
the second calculation module (103), the second calculation module (103) is used for calculating the minimum difference value of all pixel points in the pixel block in different preset directions;
the device comprises a first judgment module (104), wherein the first judgment module (104) is used for judging whether the minimum difference value in a certain preset direction is larger than a threshold value or not.
7. A device for detecting a dead pixel according to claim 5, wherein said calculation unit (200) comprises:
the second obtaining module (201), the second obtaining module (201) is configured to obtain pixels in a same row in the current frame image corresponding to a row of pixels in the previous frame for determining the suspected dead pixel, define a current pixel point as x, extend M pixels to the left side and N pixels to the right side with the current pixel point x as a center, form an inter-frame block, and form a left boundary xl and a right boundary xr in the current row between frames;
a third calculation module (202), the third calculation module (202) for calculating a minimum value, a maximum value, an average difference value and a pixel average value within an inter block;
a second judging module (203), wherein the second judging module (203) is used for judging whether the suspected dead pixel is a real dead pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210529149.3A CN115002364A (en) | 2022-05-16 | 2022-05-16 | Method and device for detecting dead pixel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210529149.3A CN115002364A (en) | 2022-05-16 | 2022-05-16 | Method and device for detecting dead pixel |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115002364A true CN115002364A (en) | 2022-09-02 |
Family
ID=83027986
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210529149.3A Pending CN115002364A (en) | 2022-05-16 | 2022-05-16 | Method and device for detecting dead pixel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115002364A (en) |
-
2022
- 2022-05-16 CN CN202210529149.3A patent/CN115002364A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785291B (en) | Lane line self-adaptive detection method | |
| WO2019148912A1 (en) | Image processing method, apparatus, electronic device and storage medium | |
| CN103763515B (en) | A kind of video abnormality detection method based on machine learning | |
| US8189913B2 (en) | Method for detecting shadow of object | |
| CN110248112B (en) | Exposure control method of image sensor | |
| US7756295B2 (en) | Change region detection device and change region detecting method | |
| CN108470356B (en) | Target object rapid ranging method based on binocular vision | |
| JP2007097178A (en) | Method for removing "red-eyes" by face detection | |
| US10951843B2 (en) | Adjusting confidence values for correcting pixel defects | |
| WO2018068300A1 (en) | Image processing method and device | |
| CN107292828B (en) | Image edge processing method and device | |
| TWI504248B (en) | Image processing apparatus and image processing method | |
| CN115002364A (en) | Method and device for detecting dead pixel | |
| TWI465699B (en) | Method of water level measurement | |
| KR101744409B1 (en) | Method and apparatus for detecting water level based on camera in low illumination environment | |
| CN116503324A (en) | Black glue module appearance detection method and system based on machine vision | |
| CN110849444A (en) | Video water level measuring method based on machine vision | |
| JP2006121642A (en) | Image detecting device, and image detection data correction method | |
| CN115802029A (en) | Image dead pixel detection method and terminal | |
| US8761516B2 (en) | Method and system for detecting the open or closed state of the eyes of a face | |
| CN114612333A (en) | Microscopic image-based microencapsulated algae cell statistical method | |
| JP2021052238A (en) | Deposit detection device and deposit detection method | |
| CN110896471A (en) | Method for detecting video transmission packet loss based on image processing | |
| CN113920065B (en) | Imaging quality evaluation method for visual detection system of industrial site | |
| CN115563998A (en) | Automatic induction triggering method based on brightness compensation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |