Simultaneous Calibration and Correction Method for Non-uniformity and Nonlinearity of Image Sensors

Abstract

This invention provides a method for simultaneous calibration and correction of non-uniformity and non-linearity of image sensors, comprising: S100, setting up a test system; S200, using the test system to test and calibrate the image sensor; S300, establishing a unified model based on the test calibration results and correcting the coefficients of the unified model. The method for simultaneous calibration and correction of non-uniformity and non-linearity of image sensors using this approach can simultaneously calibrate and correct non-uniformity and non-linearity.

Description

Technical Field

[0001] This invention relates to the field of image preprocessing technology for photoelectric imaging systems, and in particular to a method for simultaneous calibration and correction of non-uniformity and nonlinearity of image sensors. Background Technology

[0002] Non-uniformity correction techniques are divided into two categories: calibration-based correction and scene-based correction. Calibration-based correction methods are computationally simple and offer high accuracy, as correction coefficients can be quickly obtained through calibration. Scene-based correction methods achieve non-uniformity correction through image processing, independent of experimental equipment and environment. Nonlinear correction requires measuring the photoelectric response characteristic curve and using curve fitting for inverse correction.

[0003] In some scientific applications, such as those involving quantitative data inversion, consistent photoelectric response characteristics of all pixels are required, along with high linearity of the response, to minimize errors in image data acquisition. This necessitates calibration and correction of non-uniformity and non-linearity. However, traditional methods typically only address one of these two types of calibration: non-uniformity or non-linearity. They can only perform calibration and correction of non-uniformity and non-linearity individually, rather than simultaneously. This results in a cumbersome calibration process, a large number of correction coefficients, and a significant computational burden. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a method for simultaneous calibration and correction of non-uniformity and non-linearity of image sensors, which can simultaneously calibrate and correct non-uniformity and non-linearity.

[0005] To achieve the above objectives, the present invention adopts the following specific technical solution:

[0006] The image sensor non-uniformity and non-linearity synchronous calibration and correction method according to embodiments of the present invention includes:

[0007] S100, Set up the test system;

[0008] S200. Use the test system to test and calibrate the image sensor;

[0009] S300. Establish a unified model based on the test calibration results and correct the coefficients of the unified model.

[0010] According to some embodiments of the present invention, S100 includes: setting up a test system in a darkroom environment, placing the image sensor directly in front of the opening of the integrating sphere, keeping the image sensor perpendicular to the normal direction of the opening of the integrating sphere, and placing a standard reference detector at the image sensor.

[0011] According to some embodiments of the present invention, S200 includes:

[0012] S210. Turn off the integrating sphere and adjust the dark level bias of the image sensor so that the grayscale value of all pixels in the image is greater than zero.

[0013] S220. Turn on and adjust the brightness of the integrating sphere, collect image data, take the average value and store the image V, and record the output value I of the standard detector.

[0014] According to some embodiments of the present invention, the grayscale value output of the image sensor is in the range of 10% to 90% saturation value.

[0015] According to some embodiments of the present invention, S300 includes:

[0016] S310. Statistically analyze the average response of the image sensor under test;

[0017] S320. The average response curve is obtained by linear fitting of the average response using the least squares method.

[0018] S330. Calculate the target average response gray value using the average response curve;

[0019] S340. Obtain a unified model by fitting the average response gray value and correct the correction coefficient of the unified model.

[0020] According to some embodiments of the present invention, the average response of the image sensor 3 under test is statistically analyzed according to the following formula:

[0021]

[0022] Among them Let V be the mean gray level of image V, and M and N be the number of rows and columns of the image, respectively. (m,n) Represents the pixel grayscale value.

[0023] According to some embodiments of the present invention, the average response curve is linearly fitted using the least squares method to obtain:

[0024]

[0025] The coefficients k and b are calculated according to the following formula:

[0026]

[0027] Where S is the number of test calibrations.

[0028] According to some embodiments of the present invention, the coefficient b of the average response curve is set to zero, and the target average response gray value corresponding to the brightness I is calculated.

[0029] According to some embodiments of the present invention, step S340 includes: using V (m,n) As the independent variable, with For the function values, a polynomial fit is performed using the least squares method, yielding:

[0030]

[0031] in, The coefficients a0, a1, and a2 are calculated as follows to represent the average grayscale value of the response:

[0032]

[0033] According to some embodiments of the present invention, step S340 further includes: calculating correction coefficients a0(m,n), a1(m,n), and a2(m,n) pixel by pixel within the image sensor, and storing the correction coefficients.

[0034] The present invention can achieve at least the following beneficial effects: by using a unified calibration model, it realizes one-step test calibration and correction of the non-uniformity and non-linearity of the image sensor response, effectively reducing the number of calibration coefficients, thereby reducing the storage space requirement of calibration coefficients and the computational complexity of real-time correction.

[0035] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0036] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0037] Figure 1 This is a schematic diagram of a test system according to an embodiment of the present invention;

[0038] Figure 2 This is a flowchart of a method for synchronous calibration and correction of non-uniformity and non-linearity of an image sensor according to an embodiment of the invention.

[0039] Figure 3 This is a flowchart of testing and calibrating an image sensor using a testing system according to an embodiment of the invention.

[0040] Figure 4 This is a flowchart illustrating the process of establishing a unified model based on test calibration results and correcting the coefficients of the unified model according to an embodiment of the invention.

[0041] The reference numerals in the figures include:

[0042] Image acquisition computer 1, standard detector 2, image sensor 3, integrating sphere 4. Detailed Implementation

[0043] In the following description, embodiments of the invention will be described with reference to the accompanying drawings. In the description below, the same modules are denoted by the same reference numerals. Where the same reference numerals are used, their names and functions are also the same. Therefore, their detailed description will not be repeated.

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not constitute a limitation thereof.

[0045] To achieve the above objectives, the present invention adopts the following specific technical solution:

[0046] The method for simultaneous calibration and correction of non-uniformity and non-linearity of image sensor 3 according to embodiments of the present invention, such as... Figure 2 As shown, the process includes: S100, setting up a test system; S200, using the test system to test and calibrate the image sensor 3; S300, establishing a unified model and correcting the coefficients based on the test calibration results.

[0047] Traditional methods typically only address one of the two: non-uniformity correction or nonlinearity correction. In some cases, non-uniformity calibration and correction, as well as nonlinearity calibration and correction, must be performed separately. The need to calibrate non-uniformity and nonlinearity separately leads to a cumbersome calibration process. The need to correct non-uniformity and nonlinearity separately results in a large number of correction coefficients and a large amount of computation. Furthermore, the need to combine the calibration and correction results of non-uniformity and nonlinearity further complicates the calibration process, resulting in a large number of correction coefficients and a large amount of computation.

[0048] This solution achieves one-step testing, calibration, and correction of the non-uniformity and non-linearity of the image sensor 3 response through a unified calibration model. This effectively reduces the number of calibration coefficients and the amount of calibration calculation, thereby reducing the storage space requirements for calibration coefficients and the computational complexity of real-time correction. Since this solution can complete the non-uniformity and non-linearity correction in one calibration, it reduces the cumbersomeness of the calibration process.

[0049] According to some embodiments of the present invention, S100 includes: setting up a test system in a darkroom environment, placing the image sensor 3 directly in front of the opening of the integrating sphere 4, keeping the image sensor 3 perpendicular to the normal direction of the opening of the integrating sphere 4, and placing a standard reference detector at the image sensor 3.

[0050] Test System Figure 1 As shown, the image sensor 3 to be tested is placed directly in front of the opening of the integrating sphere 4. In order to improve the test results, the image sensor 3 is kept perpendicular to the normal direction of the opening of the integrating sphere 4. The standard reference detector is placed near the image sensor 3. The image acquisition computer 1 is electrically connected to the standard reference detector so that the test results can be fed back to the image acquisition computer 1.

[0051] According to some embodiments of the present invention, such as Figure 3 As shown, S200 includes:

[0052] S210. Turn off integrating sphere 4 and adjust the dark level bias of image sensor 3 so that the grayscale value of all pixels in the image is greater than zero.

[0053] S220, turn on and adjust the brightness of integrating sphere 4, collect image data, take the average value and store image V, and record the output value I of standard detector 2.

[0054] Adjust the dark level bias of image sensor 3 so that the grayscale value output by all pixels in the image is greater than zero, ensuring the accuracy of the calibration process. A slightly higher grayscale value for all pixels in the image is sufficient to achieve good results.

[0055] Adjusting the brightness of the integrating sphere 4 will change the output value I of the standard detector 2. Adjust the appropriate output value I, collect 20 image data, take the average value and store it as image V.

[0056] Understandably, the brightness of the integrating sphere 4 can be adjusted multiple times, and the number of adjustments can be determined based on the actual situation.

[0057] It is understood that the number of images collected is not limited to 20, and can be determined according to the actual situation. This embodiment of the invention does not impose any restrictions.

[0058] According to some embodiments of the present invention, the grayscale value output of the image sensor 3 is in the range of 10% to 90% saturation value.

[0059] According to some embodiments of the present invention, such as Figure 4 As shown, S300 includes:

[0060] S310. Statistically analyze the average response of the image sensor 3 under test;

[0061] S320. The average response curve is obtained by linear fitting of the average response using the least squares method.

[0062] S330. Calculate the target average response gray value using the average response curve;

[0063] S340. Obtain a unified model by fitting the average response gray value and correct the correction coefficient of the unified model.

[0064] A unified model is established based on the calibration results, and the coefficients are corrected.

[0065] S310. Statistically analyze the average response of the image sensor 3 under test.

[0066] According to some embodiments of the present invention, the average response of the image sensor 3 under test is statistically analyzed according to the following formula:

[0067]

[0068] Among them Let V be the mean gray level of image V, and M and N be the number of rows and columns of the image, respectively. (m,n) Represents the pixel grayscale value.

[0069] By acquiring the pixel grayscale values ​​V of image V (m,n) Calculate the grayscale mean of the image

[0070] M and N represent the number of rows and columns of the image, V (m,n) This represents the grayscale value of the pixels in image V at rows m and columns n, where m represents the specific row number and n represents the specific column number.

[0071] S320. The average response curve is obtained by linear fitting of the average response using the least squares method.

[0072] According to some embodiments of the present invention, the average response curve is linearly fitted using the least squares method to obtain:

[0073]

[0074] The coefficients k and b are calculated according to the following formula:

[0075]

[0076] Where S is the number of test calibrations.

[0077] The first adjustment of the integrating sphere 4 brightness was recorded as image V. 1 The output value of standard detector 2 is recorded as I1; the brightness of integrating sphere 4 is adjusted a second time, and the image recorded is V. 2 The output value of standard detector 2 is recorded as I2, and so on. The brightness of integrating sphere 4 is adjusted for the Sth time, and the recorded image is V. S The output value of the standard detector 2 is recorded as I. S Among them, the image V 1 The calculated mean gray value is denoted as From image V 2 The calculated mean gray value is denoted as And so on, from image V S The calculated mean gray value is denoted as

[0078] The average response curve was obtained by linear fitting using the least squares method. The output values ​​I1 to I2 of the standard detector 2 obtained earlier were then used. S and the calculated mean gray value to Determine the coefficients k and b for linear fitting. The average response curve obtained from linear fitting reflects the relationship between the output value and the mean gray value.

[0079] S330. Calculate the target average response gray value using the average response curve.

[0080] According to some embodiments of the present invention, the coefficient b of the average response curve is set to zero, and the target average response gray value corresponding to the brightness I is calculated.

[0081] Set the coefficient b of the average response curve to zero, and set the output values ​​I1 to I of the standard detector 2. S Substituting into the average response curve, the average response gray value is calculated. to

[0082] S340. Obtain a unified model by fitting the average response gray value and correct the correction coefficient of the unified model.

[0083] According to some embodiments of the present invention, step S340 includes: using V (m,n) As the independent variable, with For the function values, a polynomial fit is performed using the least squares method, yielding:

[0084]

[0085] in, The coefficients a0, a1, and a2 are calculated as follows to represent the average grayscale value of the response:

[0086]

[0087] Based on pixel grayscale value V 1 (m,n) To V S (m,n) and average response gray value to Calculate the coefficients a0, a1, and a2 for polynomial fitting using the least squares method, and thus obtain the polynomial with V (m,n) As the independent variable, with This is a unified model for function values.

[0088] In real-time correction, according to By using the grayscale value V of pixels in m rows and n columns (m,n) The average grayscale value of the response over m rows and n columns can then be calculated and denoted as .

[0089] According to some embodiments of the present invention, step S340 further includes: calculating correction coefficients a0(m,n), a1(m,n), and a2(m,n) pixel by pixel within the image sensor, and storing the correction coefficients. During real-time correction, based on V... (m,n) Read the corresponding correction coefficients and substitute them into the unified model. It can be calculated

[0090] Where a0(m,n), a1(m,n), and a2(m,n) refer to a0, a1, and a2 in m rows and n columns.

[0091] According to some embodiments of the present invention, the brightness of the integrating sphere 4 is adjusted 6 times to make the image data output value I1, 20 image data are collected, the average value is taken, and the image is stored as V. 1 Similarly, with image data output values ​​I2, I3, I4, I5, and I6, 20 image data points are collected, the average is taken, and the image is stored as V. 2 V 3 V 4 V 5 V 6 At this point, S is 6.

[0092] It is important to note that I2 and V 2 Correspondence; I3 and V 3 Correspondence; I4 and V 4 Corresponding to; I5 and V 5 Corresponding to; I6 and V 6 correspond.

[0093] S310. Statistically analyze the average response of the image sensor 3 under test.

[0094] According to some embodiments of the present invention, the average response of the image sensor 3 under test is statistically analyzed according to the following formula:

[0095]

[0096] Among them Let V be the mean gray level of image V, and M and N be the number of rows and columns of the image, respectively. (m,n) Represents the pixel grayscale value.

[0097] Using V 1 (m,n) get Similarly, using V 2(m,n) V 3 (m,n) V 4 (m,n) V 5 (m,n) V 6 (m,n) get The average response of the image sensor 3 under test is obtained.

[0098] S320. The average response curve is obtained by linear fitting of the average response using the least squares method.

[0099] According to some embodiments of the present invention, the average response curve is linearly fitted using the least squares method to obtain:

[0100]

[0101] Through I1, I2, I3, I4, I5, I6 and The coefficients k and b are calculated. This leads to the average response curve.

[0102] S330. Calculate the target average response gray value using the average response curve.

[0103] According to some embodiments of the present invention, the coefficient b is set to zero, and the target average response grayscale value corresponding to the brightness levels I1 to I6 is calculated. to

[0104] The linear fitting formula Set the coefficient b to zero and calculate the target average response gray value corresponding to brightness levels I1 to I6. to Among them, I1 and Correspondingly, I2 and Correspondingly, I3 and Correspondingly, I4 and Correspondingly, I5 and Correspondingly, I6 and correspond.

[0105] S340. Obtain a unified model by fitting the average response gray value and correct the correction coefficient of the unified model.

[0106] According to some embodiments of the present invention, step S340 includes: using V (m,n) As the independent variable, with For the function values, a polynomial fit is performed using the least squares method, yielding:

[0107]

[0108] in, To average the grayscale value of the response, to and to Substitute into the following formula to calculate:

[0109]

[0110] The coefficients a0, a1 and a2 are obtained.

[0111] According to some embodiments of the present invention, step S340 further includes: calculating the correction coefficients a0(m,n), a1(m,n), and a2(m,n) pixel by pixel within the image sensor, and storing the correction coefficients.

[0112] During real-time correction, according to V (m,n) Read the corresponding correction coefficients and substitute them into the unified model. It can be calculated

[0113] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0114] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

[0115] The specific embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any other corresponding changes and modifications made in accordance with the technical concept of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for simultaneous calibration and correction of non-uniformity and nonlinearity of an image sensor, characterized in that, include: S100, Set up the test system; S100 includes: setting up the test system in a darkroom environment, placing the image sensor directly in front of the opening of the integrating sphere, keeping the image sensor perpendicular to the normal direction of the opening of the integrating sphere, and placing a standard detector at the image sensor. S200. Use the test system to test and calibrate the image sensor; S200 includes: S210. Turn off the integrating sphere and adjust the dark level bias of the image sensor so that the grayscale value of all pixels in the image is greater than zero. The image sensor outputs grayscale values ​​in the range of 10% to 90% saturation. S220. Turn on and adjust the brightness of the integrating sphere, collect a preset number of image data, take the average value and store the image V, and record the output value I of the standard detector; S300. Establish a unified model based on the test calibration results and correct the coefficients of the unified model; The S300 includes: S310. Statistically analyze the average response of the image sensor under test; S320. The average response curve is obtained by linear fitting of the average response using the least squares method. S330. Calculate the target average response gray value using the average response curve; S340. Obtain a unified model by fitting the average response gray value and correct the correction coefficient of the unified model.

2. The method for simultaneous calibration and correction of image sensor non-uniformity and non-linearity as described in claim 1, characterized in that, In step S310, the average response of the sensor of the image under test is calculated according to the following formula: ； in, Let V be the mean gray level of image V, and M and N be the number of rows and columns of the image, respectively. (m, n) Represents the pixel grayscale value.

3. The method for simultaneous calibration and correction of image sensor non-uniformity and non-linearity as described in claim 2, characterized in that, The average response curve was linearly fitted using the least squares method, and the following results were obtained: ； The coefficients k and b are calculated according to the following formula: ； Where S is the number of test calibrations.

4. The method for simultaneous calibration and correction of image sensor non-uniformity and non-linearity as described in claim 3, characterized in that, Set the coefficient b of the average response curve to zero, and calculate the target average response gray value corresponding to brightness I. .

5. The method for simultaneous calibration and correction of image sensor non-uniformity and non-linearity as described in claim 4, characterized in that, Step S340 includes: using V (m, n) As the independent variable, with For the function values, a polynomial fit is performed using the least squares method, yielding: ； in, The coefficients a0, a1, and a2 are calculated as follows to represent the average grayscale value of the response: 。 6. The method for simultaneous calibration and correction of image sensor non-uniformity and non-linearity as described in claim 5, characterized in that, Step S340 further includes: calculating the correction coefficients a0(m,n), a1(m, n), and a2(m, n) for each pixel in the image sensor, and storing the correction coefficients.

Images