Device for calculating calibration data of an image

Abstract

An apparatus for calculating calibration data of an image includes a channel and region separator configured to generate a channel image by separating an original image received from an image sensor into images of respective colors and respective channels corresponding to respective pixel groups of the image sensor, and separating each channel image into a plurality of regions of interest (ROI); a channel signal extractor configured to obtain channel signals from the plurality of ROIs of the channel image; a deviation calculator configured to calculate a channel deviation between channel signals of the ROI located at the same position in the channel image having the same color; and a calibration value calculator configured to calculate a channel calibration value for each ROI based on the channel deviation and the channel signal.

Description

Technical Field

[0001] The techniques and implementations disclosed in this patent document generally relate to an apparatus for calculating calibration data. Background Art

[0002] Image sensing devices are used in electronic devices to convert optical images into electrical signals. With the development of the automotive, medical, computer and communication industries, the demand for high-performance image sensing devices is increasing in various fields such as smart phones, digital cameras, game consoles, IoT (Internet of Things), robots, security cameras and medical miniature cameras.

[0003] Image sensing devices may include a phase difference detection autofocus (PDAF) function that automatically focuses by detecting a phase difference between adjacent pixels. In order to improve the PDAF function, a structure in which multiple pixels of the same color are adjacently arranged and one microlens is applied to the multiple pixels has recently been widely used in image sensing devices. Summary of the invention

[0004] Various embodiments of the disclosed technology are directed to an apparatus capable of calculating calibration data for calibrating a signal branching phenomenon between channels caused by a manufacturing process error.

[0005] According to an embodiment of the disclosed technology, a device for calculating calibration data of an image may include: a channel and region separator, which is configured to generate a channel image by separating an original image received from an image sensor into images of individual colors and individual channels corresponding to individual pixel groups of the image, and separate each channel image into multiple regions of interest (ROIs); a channel signal extractor, which is configured to obtain channel signals from multiple ROIs of the channel image; a deviation calculator, which is configured to calculate the channel deviation between channel signals of ROIs located at the same position in the channel image having the same color; and a calibration value calculator, which is configured to calculate the channel calibration value of each ROI based on the channel deviation and the channel signal.

[0006] It is to be understood that both the foregoing general description and the following detailed description of the disclosed technology are exemplary and explanatory and are intended to provide further explanation of the disclosure as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The above and other features and advantageous aspects of the disclosed technology will become readily apparent with reference to the following detailed description when considered in conjunction with the accompanying drawings.

[0008] Figure 1 is a block diagram illustrating an example of an imaging device based on some implementations of the disclosed technology.

[0009] Figure 2Some implementations based on the disclosed technology are shown in Figure 1 A schematic diagram of an example of a pixel array included in an image sensor.

[0010] Figure 3A is a conceptual diagram illustrating an example of a method of separating an average image into multiple color images based on some implementations of the disclosed technology.

[0011] Figure 3B is a conceptual diagram illustrating an example of a method of separating individual color images into multiple channel images based on some implementations of the disclosed technology.

[0012] Figure 3C is a conceptual diagram illustrating an example of a method of separating a channel image into multiple regions of interest (ROIs) based on some implementations of the disclosed technology. DETAILED DESCRIPTION

[0013] This patent document provides implementations and examples of image processing techniques and devices for calculating calibration data required to calibrate signals of image sensors, which can be used to substantially solve one or more technical or engineering problems and alleviate limitations or disadvantages encountered in some other devices. For example, if an overlay error occurs between some structures of an image sensing device during a manufacturing process, the overlay error results in signal branching between adjacent pixels that share the same microlens for receiving incident light. Signal branching between adjacent pixels (which may be referred to as inter-channel signal branching) can cause the autofocus function in the imaging device to fail. The image processing techniques and devices disclosed in this patent document can be implemented in a manner that solves errors generated in the manufacturing process of an image sensor and provides a signal processing technique for calibrating inter-channel signal branches between channels associated with different adjacent pixels in an imaging sensor array.

[0014] Some implementations of the disclosed technology provide examples of a device capable of calculating calibration data for calibrating a signal branching phenomenon between channels caused by manufacturing process errors. The disclosed technology provides various implementations of an image sensing device that can calibrate an overlay error caused by manufacturing process errors in a signal processing step, thereby improving an autofocus function.

[0015] Reference will now be made in detail to specific embodiments, examples of which are illustrated in the accompanying drawings. Whenever possible, the same reference numerals will be used throughout the accompanying drawings to represent the same or similar parts. In the following description, detailed descriptions of related known configurations or functions contained herein will be omitted to avoid obscuring the subject matter.

[0016] Hereinafter, various embodiments will be described with reference to the accompanying drawings. However, it should be understood that the disclosed technology is not limited to a specific embodiment, but includes various modifications, equivalents and / or substitutions of the embodiments. The embodiments of the disclosed technology can provide various effects that can be directly or indirectly recognized by the disclosed technology.

[0017] Figure 1 is a block diagram illustrating an example of an imaging processing apparatus 100 for calibrating an image signal from an image sensor based on some implementations of the disclosed technology. Figure 2 Some implementations based on the disclosed technology are shown in Figure 1 A schematic diagram of an example of a pixel array included in an image sensor. FIG. 3A to FIG. 3C Some implementations based on the disclosed technology are shown in Figure 1 FIG. 1 is a diagram of an example of a method of operating channels and a region separator 120 of an exemplary imaging processing device 100 .

[0018] Reference Figure 1 , Figure 2 and FIG. 3A to FIG. 3C The calibration data calculation section of the imaging processing apparatus 100 may include an average image generator 110 , a channel and region separator 120 , a channel signal extractor 130 , a deviation calculator 140 , and a calibration value calculator 150 .

[0019] The calibration data calculation section of the imaging processing device 100 may perform calibration of data based on measurements obtained during a test process of the image sensor before the image sensor is released to the market. The calibration data calculation section of the imaging processing device 100 may detect or identify the degree of channel branching between adjacent pixels of the image sensor in a test process before the image sensor is commercialized and released, and may calculate calibration data corresponding to the degree of channel branching.

[0020] Reference Figure 2 , the pixel array of the image sensor 14 may include a plurality of pixel groups (PXG_R, PXG_Gr, PXG_Gb, PXG_B), in which four unit pixels having the same color are adjacently arranged in a (2×2) matrix. Each pixel group (PXG_R, PXG_Gr, PXG_Gb, or PXG_B) may include four channels corresponding to four adjacent unit pixels for filtering incident light to respectively detect or sense light of the same color, and the colors (R, Gr, Gb, B) corresponding to the pixel groups (PXG_R, PXG_Gr, PXG_Gb, PXG_B) may be arranged, for example, in a Bayer pattern of different color filters, in which the light intensity is 50% green, 25% red, and 25% blue. In Figure 2In the example, the numbers written in each pixel may indicate the channel of each pixel in the corresponding pixel group. Therefore, the original image generated by the image sensor 14 through shooting may include color image pixels arranged in the same structure as the above-mentioned pixel array.

[0021] The image sensor 14 can generate a raw image by photographing the light emitted from the light source 12 while changing the integration time of the photocharge generated by the incident light. Such photographing can be performed while gradually increasing the integration time (exposure time) until all pixels of the image sensor 14 become saturated. In this imaging process, the image sensor 14 can generate a plurality of raw images by performing image capture multiple times (for example, five times) for each integration time. The plurality of raw images generated by the image sensor 14 for each integration time can be input to the average image generator 110 of the imaging processing device 100.

[0022] When receiving a plurality of raw images for each integration time from the image sensor 14 , the average image generator 110 may average the plurality of raw images to generate an average image (AVG_IMG) for each integration time. The average image (AVG_IMG) generated for each integration time may be sent to the channel and region separator 120 .

[0023] The channel and region separator 120 may separate the average image (AVG_IMG) of each integration time provided from the average image generator 110 into a plurality of images of each color and each channel. Therefore, the channel and region separator 120 may generate a plurality of per-channel images (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4). Under the Bayer pattern of different color filters with light intensities of 50% green, 25% red, and 25% blue, the per-channel images R1 to R4 correspond to the color red R, the per-channel images Gr1 to Gr4 correspond to the color green Gr, the per-channel images Gb1 to Gb4 correspond to the color green Gb, and the per-channel images B1 to B4 correspond to the color blue B. In addition, the channel and region separator 120 may separate each of the per-channel images (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4) into a plurality of preset regions of interest (ROIs). Hereinafter, the average image (AVG_IMG) at each integration time will be referred to as the average image (AVG_IMG).

[0024] For example, Figure 3AAs shown, the channel and area separator 120 may first separate the average image (AVG_IMG) provided from the average image generator 110 into a plurality of images of each color (R, Gr, Gb, B) corresponding to each pixel group (PXG_R, PXG_Gr, PXG_Gb, PXG_B), and may collect images of the same color, so that four color images (R_IMG, Gr_IMG, Gb_IMG, B_IMG) may be generated. Green (Gr, Gb) may have the same color, but may be arranged at different positions within one Bayer pattern. Therefore, according to the present embodiment, two greens (Gr, Gb) from two different pixel groups may be regarded as different colors.

[0025] When collecting areas separated for each pixel group for each color (R, Gr, Gb, B) and creating color images (R_IMG, Gr_IMG, Gb_IMG, B_IMG), the areas may be arranged in the order of the corresponding pixel groups. The size of each color image (R_IMG, Gr_IMG, Gb_IMG, B_IMG) may be 1 / 4 of the size of the average image (AVG_IMG).

[0026] Reference Figure 3B , the channel and region separator 120 can separate each color image (R_IMG, Gr_IMG, Gb_IMG, B_IMG) into multiple images of each channel, collect images of the same channel, and thus generate 16 channel images (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4). Therefore, four channel images can be generated for each color image (R_IMG, Gr_IMG, Gb_IMG, B_IMG).

[0027] When collecting the regions (pixels) separated for each channel to create the channel images (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) for each channel, the corresponding regions may be arranged in the order of the corresponding color images. The size of each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) may be 1 / 16 of the size of the average image (AVG_IMG).

[0028] The channel and region separator 120 may separate each channel image (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4) into (N×N) regions of interest (ROIs) (where N is a natural number). Figure 3C An example of separating the channel image Gr1 into 12×12 ROIs is shown. Assume that Figure 3CThe channel images (R1~R4, Gr1~Gr4, Gb1~Gb4, B1~B4) shown are separated into 12×12 ROIs, and the channel and region separator 120 can separate the channel images (R1~R4, Gr1~Gr4, Gb1~Gb4, B1~B4) into 144 (=12×12) regions of interest (ROIs).

[0029] In various implementations, the number of ROIs (regions of interest) and the size of each ROI (ie, the number of pixels included in each ROI) may vary, without limitation. Figure 3C For example, the channel and region separator 120 can separate each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) into a plurality of ROIs, each ROI having a ratio Figure 3C A smaller size (where N value increases), or each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) can be separated into multiple ROIs, each ROI has a larger Figure 3C A larger size (where N value decreases). In some implementations, the channel and region separator 120 may separate each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) into M×N (M≠N, where M is a natural number) ROIs.

[0030] In the above-mentioned embodiment, each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) is separated into a predetermined number (e.g., N×N or M×N) of ROIs. In some implementations, the size of the ROI may be predetermined. When each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) is separated into a predetermined number (e.g., N×N or M×N) of ROIs, if the specification (e.g., the total number of pixels) of the product (i.e., the image sensor) changes, the number of pixels included in the ROI may also change. If the size of the ROI is predetermined, since the size of the ROI (i.e., the number of pixels) is fixed in advance, the channel and region separator 120 may separate the channel images (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) from each other based on the corresponding size of the fixed ROI.

[0031] Although in the present embodiment, the image sensor 14 generates a plurality of raw images for each integration time, other implementations are possible. For example, when the image sensor 14 outputs one raw image for each integration time, the average image generator 110 may be omitted. For example, the raw images for each integration time output from the image sensor 14 may be directly provided to the channel and region separator 120 to be separated for each color and / or for each channel.

[0032] The channel signal extractor 130 may extract (or obtain) channel signals of each ROI separated by the channel and region separator 120. For example, the channel signal extractor 130 may extract channel signals of each ROI (ROI1 to ROI144) of each channel image (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4). In some implementations, the channel signal extractor 130 may average the signal values ​​of the pixels included in the corresponding ROI among the ROIs (ROI1 to ROI144), and may extract the average signal value as the channel signal of the corresponding ROI. In this implementation, the channel signals of each ROI (ROI1 to ROI144) may be extracted for each integration time. The extracted channel signals of each ROI (e.g., the average signal value of the ROI) may be provided to the deviation calculator 140.

[0033] The deviation calculator 140 may calculate the channel deviation of each ROI using the channel signal of each ROI of each channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) provided from the channel signal extractor 130. The deviation calculator 140 may calculate the channel deviation indicating how far the channel signal of the ROI located at the same position in the four corresponding channel images (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4) of each color image (R_IMG, Gr_IMG, Gb_IMG, B_IMG) (e.g., in the four channel images having the same color) deviates from the central value of the channel signal (e.g., the average value of the channel signal).

[0034] For example, for a color image (Gr_IMG), the color image (Gr_IMG) is separated into four channel images (GR1, GR2, GR3, and GR4). Each channel image (GR1, GR2, GR3, and GR4) is then separated into ROIs. For the channel signals (Gr1_ROI1_CS, Gr2_ROI1_CS, Gr3_ROI1_CS, Gr4_ROI1_CS) of ROI1 of each channel image (GR1, GR2, GR3, and GR4), the deviation calculator 140 may calculate an average value (Gr_ROI1_CS) of the four channel signals (Gr1_ROI1_CS, Gr2_ROI1_CS, Gr3_ROI1_CS, Gr4_ROI1_CS) of each ROI (ROI1) of each of the four channel images (Gr1 to Gr4). _CSavg), each channel signal (Gr1_ROI1_CS, Gr2_ROI1_CS, Gr3_ROI1_CS, Gr4_ROI1_CS) can be divided by the average value (Gr_ROI1_CSavg) to obtain four values, and these four values ​​can be determined as the channel deviations (Gr1_ROI1_Cal_ratio, Gr2_ROI1_Cal_ratio, Gr3_ROI1_Cal_ratio, Gr4_ROI1_Cal_ratio) of the corresponding ROI (ROI1). Therefore, the channel deviation (Gr1_ROI1_Cal_ratio) may indicate the channel deviation of the ROI (ROI1) of the channel image (Gr1), the channel deviation (Gr2_ROI1_Cal_ratio) may indicate the channel deviation of the ROI (ROI1) of the channel image (Gr2), the channel deviation (Gr3_ROI1_Cal_ratio) may indicate the channel deviation of the ROI (ROI1) of the channel image (Gr3), and the channel deviation (Gr4_ROI1_Cal_ratio) may indicate the channel deviation of the ROI (ROI1) of the channel image (Gr4).

[0035] The deviation calculator 140 may calculate the channel deviations of all ROIs (ROI1 to ROI144) of the channel images (Gr1 to Gr4). In addition, the deviation calculator 140 may calculate the channel deviations of each ROI for the channel images (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4) of all color images (R_IMG, Gr_IMG, Gb_IMG, B_IMG).

[0036] In an implementation, these channel deviations may be calculated for each integration time. For example, the deviation calculator 140 may obtain the channel deviation only for the effective integration time, and may determine the average value of the channel deviation as the final channel deviation. In some implementations, the effective integration time may refer to the integration time when all signal values ​​of the channel signal of the ROI located at the same position in the channel image of the same color are within a predetermined range (e.g., 300 to 800 LSB (least significant bit)).

[0037] For example, the deviation calculator 140 may calculate the channel deviations for each integration time, may calculate the average value of each channel for only the channel deviations corresponding to the effective integration time among the calculated channel deviations, and may determine the calculated average value as the final channel deviation. If there is only one effective integration time, or if the present embodiment is implemented only for any one integration time, the channel deviation of this one integration time may become the final channel deviation.

[0038] The calibration value calculator 150 may calculate the calibration value (which may be referred to as the channel calibration value) of each ROI of the channel image (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4) by using the channel deviation (e.g., final channel deviation) calculated by the deviation calculator 140, the preset calibration constant, and the channel signal extracted by the channel signal extractor 130. For example, the calibration value calculator 150 may calculate the channel calibration value of each ROI for the channel image (R1 to R4, Gr1 to Gr4, Gb1 to Gb4, B1 to B4) by using the following Formulas 1 to 4.

[0039] The following formulas 1 to 4 may illustrate an example case of calculating channel calibration values ​​(ROI1_Gr1_Cal, ROI1_Gr2_Cal, ROI1_Gr3_Cal, ROI1_Gr4_Cal) for each ROI1 located at the same position of the corresponding channel image (Gr1 to Gr4). In the example, the channel calibration value is calculated for the ROI1 of each channel image, but other implementations are also possible. For example, the channel calibration value may be calculated for the ROI1 of each channel image.

[0040] [Formula 1]

[0041] ROI1_Gr1_Cal=Gr1_ROI1_CS×{abs(1-Gr1_ROI1_Cal_ratio)+1}×(calibration constant)

[0042] [Formula 2]

[0043] ROI1_Gr2_Cal=Gr2_ROI1_CS×{abs(1-Gr2_ROI1_Cal_ratio)+1}×(calibration constant)

[0044] [Formula 3]

[0045] ROI1_Gr3_Cal=Gr3_ROI1_CS×{abs(1-Gr3_ROI1_Cal_ratio)+1}×(calibration constant)

[0046] [Formula 4]

[0047] ROI1_Gr4_Cal=Gr4_ROI1_CS×{abs(1-Gr4_ROI1_Cal_ratio)+1}×(calibration constant)

[0048] In Formulas 1 to 4, "Gr1_ROI1_CS~Gr4_ROI1_CS" may refer to the signal value of the channel signal extracted by the channel signal extractor 130 for ROI1 of the channel image (Gr1~Gr4). "Gr1_ROI1_Cal_ratio~Gr4_ROI1_Cal_ratio" may refer to the channel deviation (e.g., final channel deviation) calculated by the deviation calculator 140 by using the signal value of the channel signal (Gr1_ROI1_CS~Gr4_ROI1_CS) and the average signal value of the signal value. When the corresponding channel deviation (Gr1_ROI1_Cal_ratio~Gr4_ROI1_Cal_ratio) is greater than 1, the calibration constant may be set to "1", and when the corresponding channel deviation is less than "1", the calibration constant may be set to "1.1". In addition, "abs" may refer to an absolute value.

[0049] If the channel signals of the ROI (ROI1) located at the same position in the channel image (Gr1~Gr4) are all the same, the channel deviations (Gr1_ROI1_Cal_ratio~Gr4_ROI1_Cal_ratio) of the ROI (ROI1) are all "1". Therefore, in the above-mentioned formulas 1 to 4, the difference between each channel deviation (Gr1_ROI1_Cal_ratio~Gr4_ROI1_Cal_ratio) and the value "1" can be the difference between each channel deviation obtained when the channel signals of the ROI (ROI1) located at the same position are equal to each other and each channel deviation calculated by the deviation calculator 140 for the corresponding ROI (ROI1) (i.e., the actual channel deviation). The calibration value calculator 150 can calculate the channel calibration value by multiplying a value proportional to the difference by the channel signal of the corresponding ROI (ROI1).

[0050] Since the channel images (Gr1 to Gr4) are obtained by separating the color image (Gr_IMG) for each channel, the region obtained by combining the ROI (ROI1) of the channel images (Gr1 to Gr4) may refer to a region located at a position corresponding to the ROI (ROI1) when the color image (Gr_IMG) is divided into (12×12) images. Therefore, the channel deviations (Gr1_ROI1_Cal_ratio, Gr2_ROI1_Cal_ratio, Gr3_ROI1_Cal_ratio, Gr4_ROI1_Cal_ratio) may be channel deviations of the ROI (ROI1) in the color image (Gr_IMG). The channel calibration values ​​of the channel deviations (Gr1_ROI1_Cal_ratio, Gr2_ROI1_Cal_ratio, Gr3_ROI1_Cal_ratio, Gr4_ROI1_Cal_ratio) may be channel calibration values ​​of the ROI (ROI1) in the color image (Gr_IMG).

[0051] Similarly, since the color images (R_IMG, Gr_IMG, Gb_IMG, B_IMG) are obtained by separating the average image (AVG_IMG) into images of respective colors, the region obtained by combining some regions corresponding to the ROI (ROI1) in the color images (R_IMG, Gr_IMG, Gb_IMG, B_IMG) may be the region located at the position corresponding to the ROI (ROI1) when the average image (AVG_IMG) is divided into (12×12) images. Therefore, the channel calibration value of the channel deviation (Gr1_ROI1_Cal_ratio, Gr2_ROI1_Cal_ratio, Gr3_ROI1_Cal_ratio, Gr4_ROI1_Cal_ratio) may become the channel calibration value of the pixel group (PXG_Gr) located at the region corresponding to the ROI (ROI1) in the average image (AVG_IMG).

[0052] By the above method, when channel calibration values ​​are calculated for each ROI (ROI1-ROI144) for all channel images (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4), channel calibration values ​​for the entire average image (AVG_IMG) can be calculated. These channel calibration values ​​can be calculated for each integration time.

[0053] In such FIG. 3A to FIG. 3CIn the above-mentioned embodiment shown, the average image (AVG_IMG) is separated into color images (R_IMG, GR_IMG, Gb_IMG, B_IMG) of each color corresponding to the pixel group. Then, each color image (R_IMG, GR_IMG, Gb_IMG, B_IMG) is separated for each channel to generate a channel image (R1-R4, Gr1-Gr4, Gb1-Gb4, B1-B4), and each channel image is separated into (12×12) ROIs. In some other implementations, the average image may be first separated into (12×12) ROIs, and the ROIs are separated for each color corresponding to each pixel group to generate a color image. Then, the color image may be separated again for each channel to generate a channel image, and a channel calibration value may also be calculated for each channel image.

[0054] The channel calibration values ​​calculated by the above-mentioned image calibration device can be stored in the memory of the image sensor manufactured by the same process as the process of the image sensor 14, so that the channel calibration values ​​can be used to calibrate the image signal in the image sensor. For example, the image sensor can separate the captured image into (12×12) ROIs (ROI1 to ROI144), and the channel values ​​(i.e., the signal values ​​of the pixel signals) of each pixel group (PXG_R, PXG_Gr, PXG_Gb, PXG_B) of each ROI can be multiplied by the channel calibration values ​​stored in the memory, so that the signal branching degree of each channel can be minimized.

[0055] As apparent from the above description, an image sensing device based on some implementations of the disclosed technology can calibrate overlay errors caused by manufacturing process errors in a signal processing step, thereby improving an auto-focusing function.

[0056] The embodiments of the disclosed technology can provide various effects that can be directly or indirectly recognized through the above-mentioned patent documents.

[0057] Although a number of exemplary embodiments have been described, it should be understood that various modifications or enhancements of the disclosed embodiments and other embodiments may be devised based on what is described and / or illustrated in this patent document.

[0058] CROSS-REFERENCE TO RELATED APPLICATIONS

[0059] This patent document claims priority to and the benefit of Korean Patent Application No. 10-2023-0087476, filed on Jul. 6, 2023, which is hereby incorporated by reference in its entirety as a part of the disclosure of this patent document.

Claims

1. A device for calculating calibration data of an image, the device comprising: a channel and region separator, which generates a channel image by separating an original image received from an image sensor into images of respective colors and respective channels corresponding to respective pixel groups of the image sensor, and separates each of the channel images into a plurality of regions of interest ROIs; a channel signal extractor, which obtains channel signals from the multiple ROIs of the channel image; A deviation calculator for calculating a channel deviation between channel signals of the ROI located at the same position in the channel image having the same color; as well as A calibration value calculator is provided for calculating a channel calibration value of each ROI based on the channel deviation and the channel signal.

2. The device according to claim 1, wherein: The channel signal extractor: The average signal value of the pixels included in each ROI is obtained as the channel signal of the corresponding ROI.

3. The device according to claim 1, wherein: The Deviation Calculator: The channel deviations of the respective colors belonging to the same pixel group are calculated.

4. The device according to claim 1, wherein: The channel and zone separators: generating a color image by separating the original image into images of respective colors corresponding to respective pixel groups; generating the channel images by separating each of the color images into images of each channel; and Each of the channel images is separated into (N×N) ROIs or (M×N) ROIs, where N and M are natural numbers.

5. The device according to claim 4, wherein: The Deviation Calculator: The channel deviations are calculated for each of the color images, and wherein corresponding channel images are separated from the same color image.

6. The device according to claim 5, wherein: The Deviation Calculator: Calculate the average value of the channel signals of the ROI located at the same position; The channel deviation of the ROI is determined by dividing the channel signal by the mean value.

7. The device according to claim 1, wherein: The channel and zone separators: generating a color image by separating the original image into images of respective colors corresponding to respective pixel groups; generating the channel images by separating each of the color images into images of each channel; and Each of the channel images is separated into a region of interest ROI having a predetermined size.

8. The device according to claim 1, wherein: The calibration value calculator: Calculating a difference between i) a channel deviation when channel signals of the ROI located at the same position are equal to each other and ii) a channel deviation calculated by the deviation calculator for the ROI; and The channel calibration value is calculated by multiplying a value proportional to the difference by the corresponding channel signal.

9. The device according to claim 1, wherein: The channel and zone separators: When the raw images of the respective integration times are received from an image sensor providing the raw images, the channel images of the respective integration times are generated.

10. The device according to claim 9, wherein: The channel and zone separators: For each integration time, a color image is generated by separating the original image into images of each color corresponding to each pixel group; generating the channel images by separating each of the color images into images of each channel; and Each of the channel images is separated into (N×N) ROIs or (M×N) ROIs, where N and M are natural numbers.

11. The device according to claim 9, wherein: The channel signal extractor: For each integration time, the average signal value of the pixels included in each ROI is obtained as the channel signal of the corresponding ROI.

12. The device according to claim 9, wherein: The Deviation Calculator: For each integration time, obtaining a channel deviation between channel signals of a region of interest ROI located at the same position for each color belonging to the same pixel group; and The value obtained when the channel deviations of the effective integration time are averaged for each channel is output as the final channel deviation.

13. The device according to claim 12, wherein: The Deviation Calculator: When the channel signals of the ROI located at the same position for the integration time are all within a predetermined range, the integration time is determined as the effective integration time.

14. The device according to claim 9, wherein: The calibration value calculator: The channel calibration value is calculated for each integration time.

15. The device according to claim 1, further comprising: An average image generator receives a plurality of raw images of respective integration times from the image sensor and generates an average image for the plurality of raw images of respective integration times.

16. The device according to claim 15, wherein: The channel and zone separators: generating the channel images by separating the average image for each integration time; and Each of the channel images is separated into the region of interest ROI.

17. The device according to claim 15, wherein: The size of each color image obtained by separating the average image for each color is one fourth the size of the average image.

18. The device according to claim 15, wherein: The size of each of the channel images obtained by separating the color image for each channel is one sixteenth of the size of the average image.

19. The device according to claim 1, wherein: The size of each of the channel images is one fourth of the size of a color image obtained by separating the original image for each color.

Images