A non-uniform dark current correction method and image sensor

Abstract

The application discloses a non-uniform dark current correction method and an image sensor. The non-uniform dark current correction method comprises the following steps: shielding the image sensor; adjusting the shielded image sensor to a first preset condition and a second preset condition respectively, and making the adjusted image sensor capture images at different temperatures respectively to obtain temperature coefficients of an active pixel array under the first preset condition and the second preset condition; adjusting the shielded image sensor to a third preset condition and a fourth preset condition respectively, and making the adjusted image sensor capture images at different gains respectively to obtain gain coefficients of the active pixel array under the third preset condition and the fourth preset condition; and correcting non-uniform dark current of the active pixel array according to the temperature coefficients of the active pixel array under the first preset condition and the second preset condition and the gain coefficients of the active pixel array under the third preset condition and the fourth preset condition.

Description

Technical Field

[0001] This invention relates to the field of image sensor technology, and in particular to a non-uniform dark current correction method and an image sensor. Background Technology

[0002] When an image sensor receives light from an external scene, it can generate a value proportional to the light intensity to form a digital image. The specific principle is that the pixel array composed of photodiodes in the image sensor can receive light signals and convert them into electric charge. The electric charge accumulates to form a voltage, and the voltage is eventually converted into a digital signal, thereby forming a digital image.

[0003] In practical applications, pixel arrays still generate electrical charges even when no incident light enters, primarily due to the presence of heat. The current generated solely by heat is called dark current, and this dark current, superimposed on the photocurrent, causes noise. Especially at high temperatures or with low signal-to-noise ratios, dark current can lead to inaccurate scene reproduction by the image sensor, affecting its imaging quality. To combat the effects of dark current, the pixel array in an image sensor is typically divided into a blacked-out pixel array and an active pixel array. The active pixel array receives light signals from the external scene; the blacked-out pixel array, blocked by an opaque metal layer, cannot receive light signals and can therefore be used to detect dark current.

[0004] If the dark current of the active pixel array and the dark current of the masked pixel array are equal, the image generated by the active pixel array can effectively filter out the influence of dark current. However, due to the different heating levels in different areas of the active pixel array, the dark current of the active pixel array will exhibit non-uniformity; hotter areas will generate larger dark currents, while cooler areas will generate smaller dark currents, causing regional differences. This means that the dark currents of the active pixel array and the masked pixel array are not completely equal. Furthermore, pixel position, gain, and exposure time also affect the dark current of the active pixel array. Therefore, using only the dark current of the masked pixel array for dark current correction of the active pixel array cannot guarantee the image quality of the image sensor. Summary of the Invention

[0005] The purpose of this invention is to provide a non-uniform dark current correction method and an image sensor, which can correct the non-uniform dark current of an active pixel array, thereby effectively improving the imaging quality of the image sensor.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution:

[0007] A non-uniform dark current correction method is used in an image sensor, wherein the pixel array in the image sensor includes an active pixel array and a black-masked pixel array; the non-uniform dark current correction method includes:

[0008] The image sensor is shielded from light;

[0009] The light-shielded image sensor is adjusted to a first preset condition and a second preset condition, and the adjusted image sensor captures images at different temperatures to obtain the temperature coefficient of the active pixel array under the first preset condition and the second preset condition; the first preset condition includes a first gain and a first exposure time, the second preset condition includes the first gain and a second exposure time, and the second exposure time is less than the first exposure time;

[0010] The image sensor, which is shielded from light, is adjusted to a third preset condition and a fourth preset condition, respectively, and the adjusted image sensor captures images at different gains to obtain the gain coefficients of the active pixel array under the third preset condition and the fourth preset condition; wherein the third preset condition includes a first temperature and a first exposure time, and the fourth preset condition includes the first temperature and a second exposure time; and

[0011] The non-uniform dark current of the active pixel array is corrected based on the temperature coefficient of the active pixel array under the first and second preset conditions, and the gain coefficient of the active pixel array under the third and fourth preset conditions.

[0012] Optionally, the different temperatures include the first temperature, a second temperature lower than the first temperature, and a plurality of intermediate temperatures between the first temperature and the second temperature;

[0013] The different gains include the first gain, a second gain less than the first gain, and a plurality of intermediate gains located between the first gain and the second gain.

[0014] Optionally, the non-uniform dark current correction method further includes: dividing the active pixel array into several pixel regions; and each pixel region includes several pixels.

[0015] Optionally, the step of obtaining the temperature coefficient of the active pixel array under the first preset condition includes:

[0016] Based on the images captured by the image sensor at different temperatures under the first preset conditions, the average pixel value of each pixel region under the first preset conditions and at the different temperatures is calculated.

[0017] Using the first gain, the first exposure time, and the set of pixel mean values ​​of all pixel regions at the first temperature as the first reference point, the pixel mean values ​​of all pixel regions under the first preset conditions and at any temperature of the different temperatures are fitted with the pixel mean values ​​of the corresponding pixel regions at the first reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array under the first preset conditions and the corresponding temperature.

[0018] Optionally, the step of obtaining the temperature coefficient of the active pixel array under the second preset condition includes:

[0019] Based on the images captured by the image sensor at different temperatures under the second preset conditions, the average pixel value of each pixel region at the second preset conditions and at the different temperatures is calculated.

[0020] Using the first gain, the second exposure time, and the set of pixel mean values ​​of all pixel regions at the first temperature as the second reference point, the pixel mean values ​​of all pixel regions at any temperature under the second preset condition are fitted with the pixel mean values ​​of the corresponding pixel regions at the second reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array under the second preset condition and the corresponding temperature.

[0021] Optionally, the step of obtaining the gain coefficient of the active pixel array under the third preset condition includes:

[0022] Based on the images captured by the image sensor under different gains and adjusted to the third preset condition, calculate the average pixel value of each pixel region under the third preset condition and the different gains;

[0023] Using the set of pixel mean values ​​of all pixel regions under the first temperature, the first exposure time, and the first gain as the third reference point, the pixel mean values ​​of all pixel regions under the third preset condition and any gain among the different gains are fitted with the pixel mean values ​​of the corresponding pixel regions in the third reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array under the third preset condition and the corresponding gain.

[0024] Optionally, the step of obtaining the gain coefficient of the active pixel array under the fourth preset condition includes:

[0025] Based on the images captured by the image sensor under different gains under the fourth preset conditions, the average pixel value of each pixel region under the fourth preset conditions and the different gains is calculated.

[0026] Using the set of pixel mean values ​​of all pixel regions under the first temperature, the second exposure time, and the first gain as the fourth reference point, the pixel mean values ​​of all pixel regions under the fourth preset condition and any gain of the different gains are fitted with the pixel mean values ​​of the corresponding pixel regions in the fourth reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array under the fourth preset condition and the corresponding gain.

[0027] Optionally, the step of correcting the non-uniform dark current of the active pixel array includes:

[0028] To enable the unshielded image sensor to capture images normally, the actual temperature, actual gain, and actual exposure time of the image sensor during operation are obtained.

[0029] Based on the temperature coefficient of the active pixel array under the first preset condition and the second preset condition, and the actual temperature, the actual temperature coefficient of the active pixel array under the first preset condition and the second preset condition is calculated using linear interpolation.

[0030] Based on the gain coefficient of the active pixel array under the third preset condition and the fourth preset condition, and the actual gain, the actual gain coefficient of the active pixel array under the third preset condition and the fourth preset condition is calculated by linear interpolation.

[0031] The product of the actual temperature coefficient of the active pixel array under the first preset condition and the actual gain coefficient of the active pixel array under the third preset condition is multiplied by the first gain, the first exposure time and the average pixel value of each pixel region under the first temperature to obtain the average pixel value of each pixel region under the actual temperature, the actual gain and the first exposure time.

[0032] The product of the actual temperature coefficient of the active pixel array under the second preset condition and the actual gain coefficient of the active pixel array under the fourth preset condition is multiplied by the first gain, the second exposure time, and the average pixel value of each pixel region under the first temperature to obtain the average pixel value of each pixel region under the actual temperature, the actual gain, and the second exposure time.

[0033] Based on the actual exposure time, linear interpolation is performed on the actual temperature, the actual gain and the pixel mean of each pixel region under the first exposure time, and the pixel mean of each pixel region under the actual temperature, the actual gain and the second exposure time to obtain the pixel mean of each pixel region under the actual exposure time, the actual temperature and the actual gain.

[0034] The pixel correction value of the active pixel array is obtained by taking the actual exposure time, the actual temperature and the average pixel value of each pixel region under the actual gain as the pixel correction value of the center pixel of the corresponding pixel region, and calculating the pixel correction value of the remaining pixels in the corresponding pixel region.

[0035] The pixel value of the active pixel array is subtracted from the pixel value of the active pixel array when the image sensor normally captures an image, in order to correct the non-uniform dark current of the active pixel array.

[0036] Optionally, bilinear interpolation is used to calculate the pixel correction values ​​for the remaining pixels within the pixel region.

[0037] Optionally, the image sensor operates normally within a preset temperature range, a preset gain range, and a preset exposure time range;

[0038] The first temperature is the maximum value of the preset temperature range, and the second temperature is the minimum value of the preset temperature range;

[0039] The first gain is the maximum value of the preset gain range, and the second gain is the minimum value of the preset gain range;

[0040] The first exposure time is the maximum value of the preset exposure time range, and the second exposure time is the minimum value of the preset exposure time range.

[0041] On the other hand, the present invention also provides a dark current correction method for an image sensor, used in an image sensor, wherein the pixel array in the image sensor includes an active pixel array and a black-masking pixel array; the dark current correction method for the image sensor includes:

[0042] The image sensor is allowed to capture images normally in order to obtain the average dark current of the blackened pixel array;

[0043] The non-uniform dark current of the active pixel array is corrected using the non-uniform dark current correction method described above; and

[0044] The dark current of the image sensor is corrected by subtracting the average dark current of the blackout pixel array from the corrected pixel values ​​of the active pixel array.

[0045] In another aspect, the present invention also provides an image sensor, including a pixel array, a processor, and a memory; the memory stores a computer program; when the computer program is executed by the processor, it implements the non-uniform dark current correction method as described above.

[0046] Optionally, the active pixel array includes a plurality of pixel regions; the memory also stores the temperature coefficient of the active pixel array under the first preset condition and the second preset condition, the gain coefficient of the active pixel array under the third preset condition and the fourth preset condition, the first gain, the first exposure time and the first temperature of each pixel region, and the first gain, the second exposure time and the first temperature of each pixel region.

[0047] Compared with the prior art, the present invention has the following advantages:

[0048] The present invention provides a non-uniform dark current correction method and an image sensor. Based on factors such as temperature, gain, and exposure time, the method calculates the temperature coefficient and gain coefficient of the active pixel array under different conditions, and uses these to correct the non-uniform dark current of the active pixel array, thereby effectively improving the imaging quality of the image sensor.

[0049] In this invention, linear regression is used to fit the temperature coefficient and gain coefficient of the active pixel array under different conditions, which can reduce the calculation cost and storage requirements of the temperature coefficient and gain coefficient. At the same time, it can eliminate some out-of-range data in certain pixel areas, making the calculated temperature coefficient and gain coefficient more reliable, thereby ensuring the correction accuracy of the non-uniform dark current of the active pixel array, and thus ensuring the imaging quality of the image sensor. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the pixel array structure in an image sensor provided by an embodiment of the present invention;

[0051] Figure 2 This is a flowchart of a non-uniform dark current correction method provided in an embodiment of the present invention;

[0052] Figure 3 This is a calculation diagram of the bilinear interpolation method in a non-uniform dark current correction method provided by an embodiment of the present invention. Detailed Implementation

[0053] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a further detailed explanation of the non-uniform dark current correction method and image sensor proposed in this invention. The advantages and features of this invention will become clearer from the following description. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, used only to facilitate and clearly illustrate the embodiments of this invention. Please refer to the accompanying drawings to make the objectives, features, and advantages of this invention more apparent and understandable. It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are only for illustrative purposes to aid those skilled in the art and are not intended to limit the implementation conditions of this invention. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to the size, without affecting the effects and objectives achieved by this invention, should still fall within the scope of the technical content disclosed in this invention.

[0054] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0055] Combined with appendix Figures 1-3 As shown, this embodiment provides a non-uniform dark current correction method for an image sensor; the pixel array 1 in the image sensor includes an active pixel array 11 and a black-masked pixel array 12. When the image sensor is working normally, the active pixel array 11 is used to receive light signals from the external scene, while the black-masked pixel array 12 does not receive light signals from the external scene.

[0056] As described in the background art, the dark current of the active pixel array 11 exhibits non-uniformity due to factors such as temperature, gain, and exposure time. Correcting the active pixel array 11 solely based on the dark current detected by the blackened pixel array 12 cannot meet the imaging quality requirements of the image sensor. This embodiment provides a non-uniform dark current correction method that calculates the temperature coefficient and gain coefficient of the active pixel array 11 under different conditions based on factors such as temperature, gain, and exposure time, and uses these to correct the non-uniform dark current of the active pixel array 11, thereby effectively improving the imaging quality of the image sensor.

[0057] Please also refer to Figure 1 and Figure 2 The non-uniform dark current correction method provided in this embodiment includes:

[0058] Step S1: Block the light from the image sensor.

[0059] Step S2: Adjust the light-shielded image sensor to a first preset condition and a second preset condition respectively, and make the adjusted image sensor (i.e., the image sensor adjusted to the first preset condition and the image sensor adjusted to the second preset condition) capture images at different temperatures respectively, so as to obtain the temperature coefficient of the active pixel array 11 under the first preset condition and the second preset condition; the first preset condition includes a first gain G1 and a first exposure time Te1, the second preset condition includes the first gain G1 and a second exposure time Te2, and the second exposure time Te2 is less than the first exposure time Te1.

[0060] Step S3: Adjust the light-shielded image sensor to the third preset condition and the fourth preset condition respectively, and make the adjusted image sensor (i.e., the image sensor adjusted to the third preset condition and the image sensor adjusted to the fourth preset condition) capture images at different gains respectively, so as to obtain the gain coefficient of the active pixel array 11 under the third preset condition and the fourth preset condition; and the third preset condition includes the first temperature T1 and the first exposure time Te1, and the fourth preset condition includes the first temperature T1 and the second exposure time Te2.

[0061] Step S4: Correct the non-uniform dark current of the active pixel array 11 according to the temperature coefficient of the active pixel array 11 under the first preset condition and the second preset condition, and the gain coefficient of the active pixel array 11 under the third preset condition and the fourth preset condition.

[0062] Specifically, in step S1, a light shield can be used to shield the image sensor so that the active pixel array 11 and the blackened pixel array 12 in the image sensor are both in a completely dark environment (i.e., without external light), thereby avoiding interference from external light on the correction of the non-uniform dark current of the active pixel array 11.

[0063] Specifically, before performing step S2, the method further includes: dividing the active pixel array 11 into several pixel regions 110; and each pixel region 110 includes several (e.g., p rows × q columns) pixels 1101, where p and q are both positive integers. Figure 1 As shown, the active pixel array 11 is divided into 9 pixel regions 110, each pixel region 110 including 9 (3 rows × 3 columns) pixels 1101, but the present invention is not limited thereto.

[0064] Specifically, in step S2, the different temperatures include the first temperature T1 and a second temperature T that is lower than the first temperature T1. n and located at the first temperature T1 and the second temperature T n Multiple intermediate temperatures (T2, T3, ..., T) n-1 ), where n > 1 and n is a positive integer. In step S3, the different gains include the first gain G1, a second gain G that is less than the first gain G1, and a third gain G that is less than the first gain G1. m and located at the first gain G1 and the second gain G m Multiple intermediate gains (G2, G3, ..., G) m-1 ), m>1 and m is a positive integer.

[0065] In one embodiment, the image sensor operates normally within a preset temperature range, a preset gain range, and a preset exposure time range. The first temperature T1 is the maximum value of the preset temperature range, and the second temperature T... n The minimum value of the preset temperature range. The first gain G1 is the maximum value of the preset gain range, and the second gain G... m The first exposure time Te1 is the maximum value of the preset exposure time range, and the second exposure time Te2 is the minimum value of the preset exposure time range. Optionally, the first temperature T1 is 80℃, and the second temperature T... n The temperature is 10℃; optionally, the first gain G1 is 24, and the second gain G m The first exposure time Te1 is 500ms, and the second exposure time Te2 is 5ms, but the present invention is not limited thereto.

[0066] Please continue to refer to this. Figure 1 and Figure 2 In step S2, the step of obtaining the temperature coefficient of the active pixel array 11 under the first preset condition includes:

[0067] Step S21: Based on the images captured by the image sensor at different temperatures under the first preset conditions, calculate the average pixel value of each pixel region 110 under the first preset conditions and at the different temperatures.

[0068] Step S22: Using the set of pixel mean values ​​of all pixel regions 110 under the first gain G1, the first exposure time Te1, and the first temperature T1 as the first reference point, the pixel mean values ​​of all pixel regions 110 under the first preset condition and at any temperature are fitted with the pixel mean values ​​of the corresponding pixel regions 110 at the first reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array 11 under the first preset condition and the corresponding temperature.

[0069] Specifically, in step S22, as shown in Table 1, the pixel mean (PT) of the pixel region 110 in the first reference point can be... 11 ,PT 12 ,PT 13 ,…,PT 1a The x-axis value is the pixel mean (e.g., PT) of the pixel region 110 under the first preset condition and at any of the different temperatures (e.g., the intermediate temperature T2). 21 ,PT 22 ,PT 23 ,…,PT 2a The vertical axis is used as the value, and a linear regression method is used to fit the most suitable slope (e.g., KT2). This slope (e.g., KT2) is multiplied by the pixel mean of the pixel region 110 in the first reference point, and the sum of the squares of the differences between this slope (e.g., KT2) and the pixel mean of the pixel region 110 under the first preset condition and the corresponding temperature (e.g., intermediate temperature T2) (e.g., (KT2 × PT)). 11 -PT 21 ) 2 +(KT2×PT 12 -PT 22 ) 2 +(KT2×PT 13 -PT 23 ) 2 +…+(KT2×PT 1a -PT 2a ) 2When the slope is at its minimum, the slope (e.g., KT2) is the temperature coefficient of the active pixel array 11 under the first preset condition and the corresponding temperature.

[0070] It is understood that 'a' represents the total number of pixel regions 110, and 'a' is a positive integer; PT 1a PT represents the average pixel value of the a-th pixel region 110 under the first gain G1, the first exposure time Te1 (i.e., the first preset condition), and the first temperature T1. 2a KT2 represents the average pixel value of the a-th pixel region 110 under the first preset condition and the intermediate temperature T2, and KT2 represents the temperature coefficient of the active pixel array 11 under the first preset condition and the intermediate temperature T2, and so on.

[0071] Table 1 Temperature coefficient of active pixel array under the first preset condition

[0072]

[0073]

[0074] Specifically, in step S2, the step of obtaining the temperature coefficient of the active pixel array 11 under the second preset condition includes:

[0075] Step S23: Based on the images captured by the image sensor at different temperatures under the second preset conditions, calculate the average pixel value of each pixel region 110 under the second preset conditions and at the different temperatures.

[0076] Step S24: Using the set of pixel average values ​​of all pixel regions 110 under the first gain G1, the second exposure time Te2, and the first temperature T1 as the second reference point, the pixel average values ​​of all pixel regions 110 under the second preset condition and at any temperature are fitted with the pixel average values ​​of the corresponding pixel regions 110 at the second reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array 11 under the second preset condition and the corresponding temperature.

[0077] Specifically, in step S24, as shown in Table 2, the pixel mean (PT) of the pixel region 110 in the second reference point can be... 11 ',PT 12 ',PT 13 ',…,PT 1a The x-axis value is the pixel mean (e.g., PT) of the pixel region 110 under the second preset condition and at any of the different temperatures (e.g., the intermediate temperature T2). 21 ',PT22 ',PT 23 ',…,PT 2a The vertical axis is represented by a value, and a linear regression method is used to fit the most suitable slope (e.g., KT2'). This slope (e.g., KT2') is the sum of the squares of the differences between the slope (e.g., KT2') multiplied by the pixel mean of the pixel region 110 in the second reference point and the pixel mean of the pixel region 110 under the second preset condition and corresponding temperature (e.g., intermediate temperature T2) (e.g., (KT2' × PT)). 11 '-PT 21 ') 2 +(KT2'×PT 12 '-PT 22 ') 2 +(KT2'×PT 13 '-PT 23 ') 2 +…+(KT2'×PT 1a '-PT 2a ') 2 When the slope is at its minimum, the slope (e.g., KT2') is the temperature coefficient of the active pixel array 11 under the second preset condition and the corresponding temperature.

[0078] Understandable, PT 1a ' represents the pixel mean of the a-th pixel region 110 under the first gain G1, the second exposure time Te2 (i.e., the second preset condition), and the first temperature T1, PT 2a ' represents the average pixel value of the a-th pixel region 110 under the second preset condition and the intermediate temperature T2, and KT2' represents the temperature coefficient of the active pixel array 11 under the second preset condition and the intermediate temperature T2, and so on.

[0079] Table 2 Temperature coefficient of active pixel array under the second preset condition

[0080]

[0081] Specifically, step S3, the step of obtaining the gain coefficient of the active pixel array 11 under the third preset condition, includes:

[0082] Step S31: Based on the images captured by the image sensor under different gains under the third preset conditions, calculate the average pixel value of each pixel region 110 under the third preset conditions and the different gains.

[0083] Step S32: Using the set of pixel mean values ​​of all pixel regions 110 under the first temperature T1, the first exposure time Te1, and the first gain G1 as the third reference point, the set of pixel mean values ​​of all pixel regions 110 under the third preset condition and any of the different gains is fitted with the pixel mean value of the corresponding pixel region 110 in the third reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array 11 under the third preset condition and the corresponding gain.

[0084] Specifically, in step S32, as shown in Table 3, the pixel mean (PG) of the pixel region 110 in the third reference point can be... 11 ,PG 12 ,PG 13 ,…,PG 1a The value of the x-axis is used as the third preset condition, and any of the different gains (e.g., the intermediate gain G) is used as the value of the x-axis. m-1 The pixel mean value (e.g., PG) of the pixel region 110 described below. (m-1)1 ,PG (m-1)2 ,PG (m-1)3 ,…,PG (m-1)a Using the values ​​of the ordinate as the vertical axis, a linear regression method is used to fit the most suitable slope (e.g., KG). m-1 ), such that the slope (e.g., KG) m-1 Multiplying the pixel mean of the pixel region 110 in the third reference point by the third preset condition and the corresponding gain (e.g., intermediate gain G) m-1 The sum of squares of the differences in the pixel mean values ​​of the pixel region 110 described below (e.g., (KG) m-1 ×PG 11 -PG (m-1)1 ) 2 +(KG m-1 ×PG 11 -PG (m-1)2 ) 2 +(KG m-1 ×PG 11 -PG (m-1)3 ) 2 +…+(KG m-1 ×PG 11 -PG (m-1)a ) 2 The slope is at its minimum (e.g., KG). m-1 ) represents the gain coefficient of the active pixel array 11 under the third preset condition and the corresponding gain.

[0085] Understandable, PG 1aPG represents the average pixel value of the a-th pixel region 110 under the first temperature T1, the first exposure time Te1 (i.e., the third preset condition), and the first gain G1. (m-1)a This indicates that under the third preset condition and with intermediate gain G m-1 The average pixel value of the a-th pixel region 110, KG m-1 This indicates the third preset condition and the intermediate gain G. m-1 The gain coefficient of the active pixel array 11 described below, and so on.

[0086] Table 3 Gain coefficients of active pixel arrays under the third preset condition

[0087]

[0088] Specifically, in step S3, the step of obtaining the gain coefficient of the active pixel array 11 under the fourth preset condition includes:

[0089] Step S33: Based on the images captured by the image sensor under different gains under the fourth preset conditions, calculate the average pixel value of each pixel region 110 under the fourth preset conditions and the different gains.

[0090] Step S34: Using the set of pixel mean values ​​of all pixel regions 110 under the first temperature T1, the second exposure time Te2, and the first gain G1 as the fourth reference point, the pixel mean values ​​of all pixel regions 110 under the fourth preset condition and any of the different gains are fitted with the pixel mean values ​​of the corresponding pixel regions 110 in the fourth reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array 11 under the fourth preset condition and the corresponding gain.

[0091] Specifically, in step S34, as shown in Table 4, the pixel mean (PG) of the pixel region 110 in the fourth reference point can be... 11 ',PG 12 ',PG 13 ',…,PG 1a The value of the x-axis is taken as '), and the fourth preset condition and any of the different gains (e.g., the intermediate gain G) are taken as '). m-1 The pixel mean value (e.g., PG) of the pixel region 110 described below. (m-1)1 ',PG (m-1)2 ',PG (m-1)3 ',…,PG (m-1)a The slope (e.g., KG) is determined by using linear regression to fit the values ​​on the ordinate. m-1 '), such that the slope (e.g., KG) m-1Multiplying the pixel mean of the pixel region 110 in the fourth reference point by the fourth preset condition and the corresponding gain (e.g., intermediate gain G) m-1 The sum of squares of the differences in the pixel mean values ​​of the pixel region 110 described below (e.g., (KG) m-1 '×PG 11 '-PG (m-1)1 ') 2 +(KG m-1 '×PG 11 '-PG (m-1)2 ') 2 +(KG m-1 '×PG 11 '-PG (m-1)3 ') 2 +…+(KG m-1 '×PG 11 '-PG (m-1)a ') 2 The slope is at its minimum (e.g., KG). m-1 ') represents the gain coefficient of the active pixel array 11 under the fourth preset condition and the corresponding gain.

[0092] Understandable, PG 1a 'PG represents the average pixel value of the a-th pixel region 110 under the first temperature T1, the second exposure time Te2 (i.e., the fourth preset condition), and the first gain G1. (m-1)a 'Indicates that under the fourth preset condition and the intermediate gain G m-1 The average pixel value of the a-th pixel region 110, KG m-1 'Indicates the fourth preset condition and intermediate gain G m-1 The gain coefficient of the active pixel array 11 described below, and so on.

[0093] Table 4 Gain coefficients of active pixel arrays under the fourth preset condition

[0094]

[0095] In this embodiment, the temperature coefficient and gain coefficient of the active pixel array 11 under different conditions are obtained by fitting using linear regression, which can reduce the calculation cost and storage requirements of the temperature coefficient and gain coefficient; at the same time, it can eliminate some outrageous data in certain pixel areas, making the calculated temperature coefficient and gain coefficient more reliable, thereby ensuring the correction accuracy of the non-uniform dark current of the active pixel array 11, and thus ensuring the imaging quality of the image sensor.

[0096] Please continue to refer to this. Figure 1 and Figure 2 Step S4 includes:

[0097] Step S41: Allow the unshielded image sensor to capture images normally to obtain the actual temperature, actual gain, and actual exposure time of the image sensor during operation.

[0098] Step S42: Based on the temperature coefficient of the active pixel array 11 under the first preset condition and the second preset condition, and the actual temperature, calculate the actual temperature coefficient of the active pixel array 11 under the first preset condition and the second preset condition using linear interpolation.

[0099] Step S43: Based on the gain coefficient of the active pixel array 11 under the third preset condition and the fourth preset condition, and the actual gain, calculate the actual gain coefficient of the active pixel array 11 under the third preset condition and the fourth preset condition respectively using linear interpolation.

[0100] Step S44: Multiply the product of the actual temperature coefficient of the active pixel array 11 under the first preset condition and the actual gain coefficient of the active pixel array 11 under the third preset condition by multiplying it by the first gain G1, the first exposure time Te1 and the first temperature T1 for each pixel region 110 to obtain the actual temperature, the actual gain and the first exposure time for each pixel region 110.

[0101] Step S45: Multiply the product of the actual temperature coefficient of the active pixel array 11 under the second preset condition and the actual gain coefficient of the active pixel array 11 under the fourth preset condition by multiplying it by the first gain G1, the second exposure time Te2 and the average pixel value of each pixel region 110 under the first temperature T1 to obtain the actual temperature, the actual gain and the average pixel value of each pixel region 110 under the second exposure time.

[0102] Step S46: Based on the actual exposure time, perform linear interpolation on the actual temperature, the actual gain, the average pixel value of each pixel region 110 under the first exposure time, and the average pixel value of each pixel region 110 under the second exposure time, to obtain the average pixel value of each pixel region 110 under the actual exposure time, the actual temperature, and the actual gain.

[0103] Step S47: Take the average pixel value of each pixel region 110 under the actual exposure time, the actual temperature and the actual gain as the pixel correction value of the center pixel of the corresponding pixel region, and calculate the pixel correction value of the remaining pixels in the corresponding pixel region to obtain the pixel correction value of the active pixel array 11.

[0104] Step S48: Subtract the pixel correction value of the active pixel array 11 from the pixel value of the active pixel array 11 when the image sensor normally captures an image, in order to correct the non-uniform dark current of the active pixel array 11.

[0105] In one embodiment, when an image is captured using the unshielded image sensor, the components in the image sensor (e.g., a temperature sensor) automatically read out the actual temperature t, actual gain g, and actual exposure time te of the image sensor.

[0106] If the actual temperature t is between the intermediate temperature T2 and the intermediate temperature T3, the actual gain g is between the intermediate gain G2 and the intermediate gain G3, and the actual exposure time te is between the first exposure time Te1 and the second exposure time Te2, then in step S42, under the first preset condition, the actual temperature coefficient Kt of the active pixel array 11 is KT2 + (actual temperature t - T2) / (T3 - T2) × (KT3 - KT2), ​​and under the second preset condition, the actual temperature coefficient Kt' of the active pixel array 11 is KT2' + (actual temperature t - T2) / (T3 - T2) × (KT3' - KT2').

[0107] In step S43, under the third preset condition, the actual gain coefficient of the active pixel array 11 is Kg = KG2 + (actual gain g - G2) / (G3 - G2) × (KG3 - KG2), and under the fourth preset condition, the actual gain coefficient of the active pixel array 11 is Kg' = KG2' + (actual gain g - G2) / (G3 - G2) × (KG3' - KG2').

[0108] In step S44, under the actual temperature t, the actual gain g, and the first exposure time Te1, the pixel average value of the first pixel region 110 is = Kt × Kg × PT. 11 The average pixel value of the second pixel region 110 is = Kt × Kg × PT 12 And so on;

[0109] In step S45, under the actual temperature t, the actual gain g, and the second exposure time Te2, the pixel average value of the first pixel region 110 is = Kt' × Kg' × PT. 11 The average pixel value of the second pixel region 110 is Kt' × Kg' × PT. 12 ', and so on;

[0110] In step S46, under the actual exposure time te, the actual temperature t, and the actual gain g, the pixel average value of the first pixel region 110 is = Kt × Kg × PT. 11 +(actual exposure time te-Te1) / (Te2-Te1)×(Kt'×Kg'×PT) 11 '-Kt×Kg×PT 11 The average pixel value of the second pixel region 110 is Kt × Kg × PT. 12 +(actual exposure time te-Te1) / (Te2-Te1)×(Kt'×Kg'×PT) 12 '-Kt×Kg×PT 12 ), and so on.

[0111] In step S47, a coordinate system is established for the active pixel array 11, and each pixel 1101 in the active pixel array 11 is regarded as a coordinate point. Based on the pixel correction value of the center pixel of each pixel region 110, the pixel correction value of the remaining pixels in each pixel region 110 is calculated using bilinear interpolation.

[0112] The calculation process of bilinear interpolation is as follows:

[0113] like Figure 3 As shown, Q12 is the known function value of the coordinate point (x1, y2), i.e., f(x1, y2); Q22 is the known function value of the coordinate point (x2, y2), i.e., f(x2, y2); Q11 is the known function value of the coordinate point (x1, y1), i.e., f(x1, y1); and Q21 is the function value of the coordinate point (x2, y1), i.e., f(x2, y1).

[0114] P is the function value to be determined at the coordinate point (x,y), i.e., f(x,y) is calculated using bilinear interpolation.

[0115] First, perform two linear interpolations in the x-direction to obtain f(x,y1)=(x2-x) / (x2-x1)×Q11+(x-x1) / (x2-x1)×Q21, f(x,y2)=(x2-x) / (x2-x1)×Q12+(x-x1) / (x2-x1)×Q22;

[0116] Then, by performing a linear interpolation in the y direction, we can obtain f(x,y)=(y2-y) / (y2-y1)×f(x,y1)+(y-y1) / (y2-y1)×f(x,y2).

[0117] In step S48, the pixel value of each pixel in the active pixel array 11 when the image sensor normally captures an image is subtracted from the pixel correction value of the corresponding pixel in the active pixel array 11, thereby correcting the non-uniform dark current of the active pixel array 11 and improving the imaging quality of the image sensor.

[0118] On the other hand, this embodiment also provides a dark current correction method for an image sensor, used in an image sensor where the pixel array 1 includes an active pixel array 11 and a blackout pixel array 12. The dark current correction method for the image sensor includes: enabling the image sensor to capture an image normally to obtain the average dark current of the blackout pixel array 12; correcting the non-uniform dark current of the active pixel array 11 using the non-uniform dark current correction method described above; and subtracting the average dark current of the blackout pixel array 12 from the corrected pixel values ​​of the active pixel array 11 to correct the dark current of the image sensor.

[0119] Specifically, in this embodiment, the dark current correction of the image sensor includes two corrections: non-uniform dark current correction of the active pixel array 11 and dark current correction of the dark pixel array 12. These two corrections significantly improve the imaging quality of the image sensor, thereby meeting image quality requirements. It is understood that the calculation method for the average dark current of the dark pixel array 12 in the dark current correction is similar to existing technologies and will not be elaborated further here.

[0120] In another aspect, this embodiment also provides an image sensor, including a pixel array 1, a processor, and a memory; the memory stores a computer program; when the computer program is executed by the processor, it implements the non-uniform dark current correction method as described above.

[0121] Optionally, the active pixel array 11 includes a plurality of pixel regions 110; the memory also stores the temperature coefficient of the active pixel array 11 under the first preset condition and the second preset condition, the gain coefficient of the active pixel array 11 under the third preset condition and the fourth preset condition, the first gain, the first exposure time and the first temperature, and the first gain, the second exposure time and the first temperature, and the first temperature, and the first gain, the second exposure time and the first temperature, and the first temperature, and the first pixel gain, the second exposure time and the first temperature, and the first temperature, respectively.

[0122] In summary, this embodiment provides a non-uniform dark current correction method and image sensor. Based on factors such as temperature, gain, and exposure time, it calculates the temperature coefficient and gain coefficient of the active pixel array under different conditions, and uses these to correct the non-uniform dark current of the active pixel array, thereby effectively improving the imaging quality of the image sensor. This embodiment uses linear regression to fit the temperature coefficient and gain coefficient of the active pixel array under different conditions, which reduces the computational cost and storage requirements for the temperature coefficient and gain coefficient. Simultaneously, it can eliminate data from certain pixel regions that are too extreme, making the calculated temperature coefficient and gain coefficient more reliable, thus ensuring the accuracy of the correction of the non-uniform dark current of the active pixel array, and ultimately guaranteeing the imaging quality of the image sensor.

[0123] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A non-uniform dark current correction method for an image sensor, wherein the pixel array in the image sensor comprises an active pixel array and a black-masking pixel array; characterized in that, The non-uniform dark current correction method includes: The image sensor is shielded from light; The light-shielded image sensor is adjusted to a first preset condition and a second preset condition, and the adjusted image sensor captures images at different temperatures to obtain the temperature coefficient of the active pixel array under the first preset condition and the second preset condition; the first preset condition includes a first gain and a first exposure time, the second preset condition includes the first gain and a second exposure time, and the second exposure time is less than the first exposure time; The image sensor, which is shielded from light, is adjusted to a third preset condition and a fourth preset condition, respectively, and the adjusted image sensor captures images at different gains to obtain the gain coefficients of the active pixel array under the third preset condition and the fourth preset condition; wherein the third preset condition includes a first temperature and a first exposure time, and the fourth preset condition includes the first temperature and a second exposure time; and The non-uniform dark current of the active pixel array is corrected based on the temperature coefficient of the active pixel array under the first and second preset conditions, and the gain coefficient of the active pixel array under the third and fourth preset conditions.

2. The non-uniform dark current correction method as described in claim 1, characterized in that, The different temperatures include the first temperature, a second temperature lower than the first temperature, and a plurality of intermediate temperatures between the first temperature and the second temperature; The different gains include the first gain, a second gain less than the first gain, and a plurality of intermediate gains located between the first gain and the second gain.

3. The non-uniform dark current correction method as described in claim 1, characterized in that, Also includes: The active pixel array is divided into several pixel regions; and each pixel region includes several pixels.

4. The non-uniform dark current correction method as described in claim 3, characterized in that, The step of obtaining the temperature coefficient of the active pixel array under the first preset condition includes: Based on the images captured by the image sensor at different temperatures under the first preset conditions, the average pixel value of each pixel region under the first preset conditions and at the different temperatures is calculated. Using the first gain, the first exposure time, and the set of pixel mean values ​​of all pixel regions at the first temperature as the first reference point, the pixel mean values ​​of all pixel regions under the first preset conditions and at any temperature of the different temperatures are fitted with the pixel mean values ​​of the corresponding pixel regions at the first reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array under the first preset conditions and the corresponding temperature.

5. The non-uniform dark current correction method as described in claim 3, characterized in that, The step of obtaining the temperature coefficient of the active pixel array under the second preset condition includes: Based on the images captured by the image sensor at different temperatures under the second preset conditions, the average pixel value of each pixel region at the second preset conditions and at the different temperatures is calculated. Using the first gain, the second exposure time, and the set of pixel mean values ​​of all pixel regions at the first temperature as the second reference point, the pixel mean values ​​of all pixel regions at any temperature under the second preset condition are fitted with the pixel mean values ​​of the corresponding pixel regions at the second reference point using a linear regression method, and the slope obtained by fitting is the temperature coefficient of the active pixel array under the second preset condition and the corresponding temperature.

6. The non-uniform dark current correction method as described in claim 3, characterized in that, The steps for obtaining the gain coefficient of the active pixel array under the third preset condition include: Based on the images captured by the image sensor under different gains and adjusted to the third preset condition, calculate the average pixel value of each pixel region under the third preset condition and the different gains; Using the set of pixel mean values ​​of all pixel regions under the first temperature, the first exposure time, and the first gain as the third reference point, the pixel mean values ​​of all pixel regions under the third preset condition and any gain among the different gains are fitted with the pixel mean values ​​of the corresponding pixel regions in the third reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array under the third preset condition and the corresponding gain.

7. The non-uniform dark current correction method as described in claim 3, characterized in that, The step of obtaining the gain coefficient of the active pixel array under the fourth preset condition includes: Based on the images captured by the image sensor under different gains under the fourth preset conditions, the average pixel value of each pixel region under the fourth preset conditions and the different gains is calculated. Using the set of pixel mean values ​​of all pixel regions under the first temperature, the second exposure time, and the first gain as the fourth reference point, the pixel mean values ​​of all pixel regions under the fourth preset condition and any gain of the different gains are fitted with the pixel mean values ​​of the corresponding pixel regions in the fourth reference point using a linear regression method, and the slope obtained by fitting is the gain coefficient of the active pixel array under the fourth preset condition and the corresponding gain.

8. The non-uniform dark current correction method as described in claim 3, characterized in that, The steps for correcting the non-uniform dark current of the active pixel array include: To enable the unshielded image sensor to capture images normally, the actual temperature, actual gain, and actual exposure time of the image sensor during operation are obtained. Based on the temperature coefficient of the active pixel array under the first preset condition and the second preset condition, and the actual temperature, the actual temperature coefficient of the active pixel array under the first preset condition and the second preset condition is calculated using linear interpolation. Based on the gain coefficient of the active pixel array under the third preset condition and the fourth preset condition, and the actual gain, the actual gain coefficient of the active pixel array under the third preset condition and the fourth preset condition is calculated by linear interpolation. The product of the actual temperature coefficient of the active pixel array under the first preset condition and the actual gain coefficient of the active pixel array under the third preset condition is multiplied by the first gain, the first exposure time and the average pixel value of each pixel region under the first temperature to obtain the average pixel value of each pixel region under the actual temperature, the actual gain and the first exposure time. The product of the actual temperature coefficient of the active pixel array under the second preset condition and the actual gain coefficient of the active pixel array under the fourth preset condition is multiplied by the first gain, the second exposure time, and the average pixel value of each pixel region under the first temperature to obtain the average pixel value of each pixel region under the actual temperature, the actual gain, and the second exposure time. Based on the actual exposure time, linear interpolation is performed on the actual temperature, the actual gain and the pixel mean of each pixel region under the first exposure time, and the pixel mean of each pixel region under the actual temperature, the actual gain and the second exposure time to obtain the pixel mean of each pixel region under the actual exposure time, the actual temperature and the actual gain. The pixel correction value of the active pixel array is obtained by taking the actual exposure time, the actual temperature and the average pixel value of each pixel region under the actual gain as the pixel correction value of the center pixel of the corresponding pixel region, and calculating the pixel correction value of the remaining pixels in the corresponding pixel region. The pixel value of the active pixel array is subtracted from the pixel value of the active pixel array when the image sensor normally captures an image, in order to correct the non-uniform dark current of the active pixel array.

9. The non-uniform dark current correction method as described in claim 8, characterized in that, The pixel correction values ​​of the remaining pixels within the pixel region are calculated using bilinear interpolation.

10. The non-uniform dark current correction method as described in claim 2, characterized in that, The image sensor operates normally within a preset temperature range, a preset gain range, and a preset exposure time range. The first temperature is the maximum value of the preset temperature range, and the second temperature is the minimum value of the preset temperature range; The first gain is the maximum value of the preset gain range, and the second gain is the minimum value of the preset gain range; The first exposure time is the maximum value of the preset exposure time range, and the second exposure time is the minimum value of the preset exposure time range.

11. A dark current correction method for an image sensor, used in an image sensor, wherein the pixel array in the image sensor includes an active pixel array and a black-masking pixel array; characterized in that, The dark current correction method for the image sensor includes: The image sensor is allowed to capture images normally in order to obtain the average dark current of the blackened pixel array; The non-uniform dark current of the active pixel array is corrected using the non-uniform dark current correction method as described in any one of claims 1 to 10; and The dark current of the image sensor is corrected by subtracting the average dark current of the blackout pixel array from the corrected pixel values ​​of the active pixel array.

12. An image sensor, characterized in that, It includes a pixel array, a processor, and a memory; the memory stores a computer program; when the computer program is executed by the processor, it implements the method as described in any one of claims 1 to 10.

13. The image sensor as claimed in claim 12, characterized in that, The active pixel array includes several pixel regions; the memory also stores the temperature coefficient of the active pixel array under the first preset condition and the second preset condition, the gain coefficient of the active pixel array under the third preset condition and the fourth preset condition, the first gain, the first exposure time and the first temperature of each pixel region, and the first gain, the second exposure time and the first temperature of each pixel region.

Images