Method and system for reducing image sensor column noise

Abstract

The present invention provides a method and system for reducing the column noise of an image sensor, the method comprising: reading the dark pixels in each frame of an image in sequence to obtain the dark pixel data in each frame of an image; and reading the initial effective pixel data in each frame of an image; calculating the dark pixel data in sequence to obtain the mean value of the dark pixels in each column and the overall mean value of the dark pixels in each frame of an image; obtaining the correction value of each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels; calculating the correction value of each column in each frame of an image with the corresponding initial effective pixel data to obtain the target effective pixel data. The present invention pre-processes the dark pixel data, calculates the correction value of each column from the dark pixel data, then inverts the correction value of each column, and performs a sum operation with the initial effective pixel data to obtain the target effective pixel data. By subtracting the mismatch value, the column noise is effectively suppressed, and the image quality is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated circuits, and in particular to a method and system for reducing image sensor column noise. Background Art

[0002] Complementary Metal-Oxide-Semiconductor (CMOS) image sensors have been widely used in imaging fields such as video, monitoring, industrial manufacturing, automobiles, and home appliances. Its mainstream readout circuit structure is a readout circuit based on column-level analog-to-digital converters (ADCs), which ensures sufficient conversion accuracy and speed under reasonable power consumption. However, due to factors such as processing technology, ADCs in different columns will produce column-to-column deviations, which are fixed. When the system gain is large, the difference is magnified, and fixed column noise will be clearly seen on the image, greatly affecting the image quality.

[0003] Column noise is also affected by temperature, process angle, exposure time, etc. To ensure that column noise is suppressed under different environmental conditions, it is generally necessary to introduce a more complex algorithm to solve the problem in the background data processing, and store a large amount of multi-frame image data. By storing multiple frames of images and taking the average deviation, and then subtracting this deviation from the existing image, the synthesized image will not see the column noise caused by this deviation. However, this requires the system to provide a memory chip for the CMOS image sensor to store multiple frames of image data, which increases the system cost a lot, and this method is not applicable in applications with high output frame rates. Summary of the invention

[0004] The object of the present invention is to provide a method and system for reducing the column noise of an image sensor, which can effectively eliminate the influence of the column noise and improve the image quality.

[0005] To achieve the above-mentioned purpose, in a first aspect, the present invention provides a method for reducing the column noise of an image sensor, the method comprising: reading the dark pixels in each frame of image in turn to obtain the dark pixel data in each frame of image; and reading the initial effective pixel data in each frame of image; calculating the dark pixel data in turn to obtain the mean value of the dark pixels in each column and the overall mean value of the dark pixels in each frame of image; obtaining a correction value for each column based on the mean value of the dark pixels in each column and the overall mean value of the dark pixels; calculating the correction value of each column in each frame of image with the corresponding initial effective pixel data to obtain the target effective pixel data.

[0006] The method for reducing the image sensor column noise provided by the present invention has the beneficial effect of: processing the dark pixel data, calculating the correction value of each column from the dark pixel data, and then calculating the correction value of each column with the initial effective pixel data to obtain the target effective pixel data. By subtracting the mismatch value, the column noise is effectively suppressed, so that the image quality is improved.

[0007] Optionally, the dark pixel data includes N columns of dark pixels, where N is a natural number;

[0008] The calculating the dark pixel data to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels includes:

[0009] Calculate the column mean of the N columns of dark pixels to obtain the mean of the dark pixels in each column; and

[0010] An overall mean value of the N columns of dark pixels is calculated to obtain an overall mean value of the dark pixels.

[0011] Optionally, obtaining the correction value of each column according to the mean value of the dark pixels of each column and the mean value of the dark pixels as a whole includes: calculating the difference between the mean value of the dark pixels of each column and the mean value of the dark pixels as a whole to obtain the correction value of each column; after obtaining the correction value of each column, the correction value of each column is stored in a register. The beneficial effect is that the correction value of each column is easy to apply.

[0012] Optionally, the dark pixels in each frame of the image are read in sequence to obtain the dark pixel data in each frame of the image: after the pixel array performs pixel photosensitization, the obtained photocharge is converted into an analog voltage; the analog voltage is converted into an analog-to-digital value to obtain the dark pixel data; wherein the pixel array is composed of a plurality of pixel units.

[0013] Optionally, the correction value of each column in each frame image is calculated with the corresponding initial effective pixel data to obtain target effective pixel data, including: inverting the correction value of each column and summing it with the initial effective pixel data to obtain the target effective pixel data; or, performing a difference calculation between the correction value of each column and the initial effective pixel data to obtain the target effective pixel data.

[0014] In a second aspect, an embodiment of the present invention provides a system for reducing image sensor column noise, comprising: a reading unit, configured to sequentially read dark pixels in each frame of image to obtain dark pixel data in each frame of image; and read initial valid pixel data in each frame of image;

[0015] A data processing unit, used for calculating the dark pixel data in sequence to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels in each frame of image;

[0016] The data processing unit is further used to obtain a correction value for each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels;

[0017] The reading unit is further used to read the initial valid pixel data in each frame of image;

[0018] The correction unit inverts the correction value of each column in each frame of the image, performs a sum operation with the corresponding initial effective pixel data, and obtains target effective pixel data.

[0019] The beneficial effect of the present invention is that by processing dark pixel data and calculating the correction value of each column from the dark pixel data, the correction value of each column is calculated with the initial effective pixel data to obtain the target effective pixel data. By subtracting the mismatch value, the column noise is effectively suppressed, so that the image quality is improved.

[0020] Optionally, the dark pixel data includes N columns of dark pixels, where N is a natural number;

[0021] The data processing unit is used to perform column mean calculation on the N columns of dark pixels to obtain the mean of the dark pixels in each column; and perform overall mean calculation on the N columns of dark pixels to obtain the overall mean of the dark pixels.

[0022] Optionally, the data processing unit is used to obtain the correction value of each column based on the mean value of the dark pixels in each column and the overall mean value of the dark pixels, and is also used to: calculate the difference between the mean value of the dark pixels in each column and the overall mean value of the dark pixels to obtain the correction value of each column; after obtaining the correction value of each column, it also includes: storing the correction value of each column in a register.

[0023] Optionally, after a plurality of pixel units perform pixel light sensing, the obtained photocharge is converted into an analog voltage amount;

[0024] The column analog-to-digital converter corresponding to the pixel unit performs analog-to-digital conversion to obtain the dark pixel data.

[0025] Optionally, the correction unit negates the correction value of each column in each frame image, performs a sum operation with the corresponding initial effective pixel data, and obtains the target effective pixel data, including: negating the correction value of each column, performing a sum operation with the initial effective pixel data, and obtaining the target effective pixel data; or

[0026] The correction value of each column is subjected to difference calculation with the initial effective pixel data to obtain the target effective pixel data. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1A flow chart of a method for reducing image sensor column noise according to an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of reducing the overall circuit architecture of an image sensor according to an embodiment of the present invention;

[0029] Figure 3 A schematic diagram of the structure of a system for reducing image sensor column noise according to an embodiment of the present invention;

[0030] Figure 4 A detailed diagram of a system for reducing image sensor column noise according to an embodiment of the present invention;

[0031] Figure 5 A schematic diagram of an output image when the correction logic of an embodiment provided by the present invention is not used;

[0032] Figure 6 A schematic diagram of an output image when the correction logic provided in an embodiment of the present invention is used. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0034] In view of the problems existing in the prior art, an embodiment of the present invention provides a method for reducing the noise of an image sensor column. Figure 1 As shown, the method includes:

[0035] S101: reading dark pixels in each frame of image in sequence to obtain dark pixel data in each frame of image; and reading initial valid pixel data in each frame of image.

[0036] Before this step, after the pixel array performs pixel photosensitization, the obtained photocharge is converted into an analog voltage, and then the analog voltage is converted into a corresponding digital quantity by analog-to-digital conversion, that is, the dark pixel data is obtained. It should be noted that the pixel array is composed of a plurality of columns of pixel units, each column of pixel units corresponds to a column analog-to-digital converter, and the dark pixel data is finally obtained and temporarily stored in the column storage unit for subsequent processing.

[0037] It should be noted that the initial valid pixel data includes N columns of initial valid pixels.

[0038] S102: Calculate the dark pixel data in sequence to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels in each frame of image.

[0039] In this step, the dark pixel data includes N columns of dark pixels, where N is a natural number. The mean values ​​of the dark pixels in each column are obtained by performing column mean calculation on the N columns of dark pixels; and the overall mean value of the dark pixels is obtained by performing overall mean calculation on the N columns of dark pixels.

[0040] S103: Obtain correction values ​​for each column according to the mean values ​​of the dark pixels in each column and the overall mean value of the dark pixels.

[0041] In this embodiment, the correction value of each column is obtained by performing a difference calculation between the mean value of the dark pixels in each column and the mean value of the dark pixels as a whole, and the correction value of each column is stored in a register for easy access. In some embodiments, the mean value of the dark pixels in each column may be inverted and then summed with the mean value of the dark pixels as a whole to obtain the correction value of each column.

[0042] S104: Calculate the correction value of each column in each frame of image and the corresponding initial effective pixel data to obtain target effective pixel data.

[0043] In this step, the correction value of each column is inverted from the register, and then summed with the corresponding N columns of initial effective pixels to obtain target effective pixel data. Alternatively, the correction value of each column is directly differenced with the initial effective pixel data to obtain the target effective pixel data.

[0044] In this embodiment, reference Figure 2As shown, after the pixel array is photosensitive, the obtained photocharge is converted into an analog voltage, and then the analog voltage is converted into a corresponding digital quantity by an analog-to-digital converter (ADC), and the dark pixel data is obtained, and the dark pixel data is sequentially stored in the corresponding storage unit (ME), and then the mean value of each column of dark pixels and the overall mean value of dark pixels are obtained through data processing, and the mean value of each column of dark pixels is subtracted from the overall mean value of dark pixels to obtain the correction value of each column, which will be stored in the register for correction of each column in the correction unit (CR). It should be noted that in other embodiments, a device with a storage function can be used instead of the register to realize the storage of the correction value, which will not be elaborated here. When each frame of the image is reading out the effective pixel data, the correction value of each column in the register is inverted and summed, so that the mismatch value (Mismatch value) of the column is subtracted, thereby effectively suppressing the column noise and improving the image quality.

[0045] In another embodiment of the present invention, a system for reducing image sensor column noise is provided, referring to Figure 3 As shown, the system includes: a reading unit 310, which is used to read the dark pixels in each frame of image in turn to obtain the dark pixel data in each frame of image; and read the initial effective pixel data in each frame of image. A data processing unit 320 is used to calculate the dark pixel data in turn to obtain the mean value of the dark pixels in each column and the overall mean value of the dark pixels in each frame of image. The data processing unit 320 is also used to calculate the correction value of each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels. The reading unit 310 is also used to read the initial effective pixel data in each frame of image. The correction unit 330 inverts the correction value of each column in each frame of image, and performs a sum operation with the corresponding initial effective pixel data to obtain the target effective pixel data.

[0046] In this embodiment, by processing the dark pixel data and calculating the correction value of each column from the dark pixel data, the correction value of each column is negated and then operated with the initial effective pixel data of the corresponding column to obtain the target effective pixel data. By subtracting the mismatch value, the column noise is effectively suppressed, so that the image quality is improved.

[0047] Specifically, refer to Figure 4As shown, the data processing unit 320 includes a column mean operation module 321, an overall mean operation module 322, a subtraction operation module 323 and a background data processing module 324. The column mean operation module 321 and the overall mean operation module 322 are both electrically connected to the subtraction operation module 323. The correction unit 330 includes a register 331, a negation logic module 332 and an addition operation module 330. The subtraction operation module 323 is electrically connected to the register 331. The register 331 is electrically connected to the negation logic module 332 and the addition operation module 330 in sequence. The dark pixel data includes N columns of dark pixels, where N is a natural number. The column mean operation module 321 in the data processing unit 320 is used to perform column mean calculation on the N columns of dark pixels to obtain the mean of each column of dark pixels. The overall mean operation module 322 performs overall mean calculation on the N columns of dark pixels to obtain the overall mean of the dark pixels. The subtraction module 323 calculates the difference between the mean value of the dark pixels in each column and the overall mean value of the dark pixels to obtain the correction value of each column, and stores the correction value in the register 331 in the correction unit 330. After the reading unit 330 reads the initial effective pixel data in each frame of the image, the inversion logic module 332 inverts the correction value. The addition module 333 performs a sum operation on the inverted correction value and the initial effective pixel data to obtain the target effective pixel data. The target effective pixel data is used for processing by the background data processing module 324.

[0048] refer to Figure 5 and Figure 6 As shown, under the same other preconditions, Figure 5 To correct the output image when the logic is not used, Figure 6 This is the output image when the correction logic is working. Through comparison, it is found that the correction logic can effectively eliminate the influence of column noise and improve the image quality.

[0049] The above is only a specific implementation of the embodiment of the present application, but the protection scope of the embodiment of the present application is not limited thereto, and any changes or replacements within the technical scope disclosed in the embodiment of the present application should be included in the protection scope of the embodiment of the present application. Therefore, the protection scope of the embodiment of the present application should be based on the protection scope of the claims.

Claims

1. A method for reducing image sensor column noise, characterized in that: include: Read the dark pixels in each frame of image in sequence to obtain the dark pixel data in each frame of image; and reading the initial effective pixel data in each frame of the image; wherein the dark pixels and the initial effective pixels are arranged in N columns, where N is a natural number; Calculating the dark pixel data in sequence to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels in each frame of image; Obtaining a correction value for each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels; The correction value of each column in each frame image is calculated with the corresponding initial effective pixel data to obtain target effective pixel data, wherein the correction value of each column is inverted and summed with the initial effective pixel data to obtain the target effective pixel data; or, the correction value of each column is differenced with the initial effective pixel data to obtain the target effective pixel data.

2. The method according to claim 1, characterized in that The calculating the dark pixel data to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels includes: Calculate the column mean of the N columns of dark pixels to obtain the mean of the dark pixels in each column; and An overall mean value of the N columns of dark pixels is calculated to obtain an overall mean value of the dark pixels.

3. The method according to claim 1 or 2, characterized in that: The step of obtaining a correction value of each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels comprises: Calculate the difference between the mean value of the dark pixels in each column and the overall mean value of the dark pixels to obtain the correction value of each column; After obtaining the correction value of each column, the method includes: The correction value of each column is stored in a register.

4. The method according to claim 1, characterized in that The step of sequentially reading dark pixels in each frame of image to obtain dark pixel data in each frame of image includes: After the pixel array performs pixel light sensing, the obtained photocharge is converted into an analog voltage quantity; Performing analog-to-digital conversion on the analog voltage to obtain the dark pixel data; Wherein, the pixel array is composed of a plurality of pixel units.

5. A system for reducing image sensor column noise, characterized in that: include: A reading unit, used for sequentially reading dark pixels in each frame of image to obtain dark pixel data in each frame of image; And read the initial valid pixel data in each frame of image; A data processing unit, used to calculate the dark pixel data in sequence to obtain the mean value of dark pixels in each column and the overall mean value of dark pixels in each frame of image; The data processing unit is further used to obtain a correction value for each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels; The correction unit inverts the correction value of each column in each frame of the image, performs a sum operation with the corresponding initial effective pixel data, and obtains target effective pixel data.

6. The system according to claim 5, characterized in that The dark pixel data includes N columns of dark pixels, where N is a natural number; The data processing unit is used to perform column mean calculation on the N columns of dark pixels to obtain the mean of the dark pixels in each column; and perform overall mean calculation on the N columns of dark pixels to obtain the overall mean of the dark pixels.

7. The system according to claim 5 or 6, characterized in that: The data processing unit is used to obtain the correction value of each column according to the mean value of the dark pixels in each column and the overall mean value of the dark pixels, and is also used to: Calculate the difference between the mean value of the dark pixels in each column and the overall mean value of the dark pixels to obtain the correction value of each column; After obtaining the correction value of each column, the method further includes: storing the correction value of each column in a register.

8. The system according to claim 5, characterized in that After a number of pixel units perform pixel light sensing, the obtained photocharge is converted into an analog voltage; The column analog-to-digital converter corresponding to the pixel unit performs analog-to-digital conversion to obtain the dark pixel data.

9. The system according to claim 5, characterized in that The correction unit inverts the correction value of each column in each frame image, performs a sum operation with the corresponding initial effective pixel data, and obtains target effective pixel data, including: The correction value of each column is inverted, and the correction value is summed with the initial effective pixel data to obtain the target effective pixel data; or, The correction value of each column is subjected to difference calculation with the initial effective pixel data to obtain the target effective pixel data.

10. The system according to claim 5, characterized in that The data processing unit includes a column mean calculation module, an overall mean calculation module and a subtraction calculation module; The column mean calculation module and the overall mean calculation module are both electrically connected to the subtraction calculation module, and the column mean calculation module and the overall mean calculation module are both electrically connected to the reading unit.

11. The system according to claim 10, characterized in that The correction unit includes a register, a negation logic module and an addition operation module; The register is electrically connected to the inversion logic module and the addition operation module in sequence; The addition module is also electrically connected to the reading module.

12. The system according to claim 11, characterized in that The data processing unit further comprises a background data processing module, and the background data processing module is electrically connected to the addition operation module.

Images