An improved adaptive shutter correction method for infrared images

Abstract

This invention addresses the technical problem that existing technologies using shutter correction to improve the uniformity of infrared detection imaging suffer from reduced image uniformity due to environmental factors and lens limitations, leading to additional non-uniform noise and impacting image quality. The invention provides an improved adaptive shutter correction method for infrared images. This method includes the following steps: Step 1) Calibrating shutter parameters: Acquiring two blackbody images at high and low temperatures, calculating the gain coefficient; calculating non-uniformity calibration correction parameters; Step 2) Shutter correction: Closing the shutter, acquiring the image to be corrected as the current shutter image data; calculating the correlation between the current shutter image data and the low-temperature blackbody image data, and correcting the non-uniformity calibration correction parameters based on this correlation; traversing the pixels in the image to be corrected, and combining the corrected non-uniformity correction parameters, calculating the corrected image data.

Description

Technical Field

[0001] This invention relates to the field of infrared detection and imaging technology, and specifically to an improved adaptive shutter correction method for infrared images. Background Technology

[0002] During infrared detection and imaging, infrared detectors are affected by various environmental and internal factors, leading to an increasing trend in the non-uniformity of infrared images over time, severely impacting the image quality. To improve the stability of image quality over time, shutter correction is needed to enhance image uniformity. Currently, two-point correction is employed for non-uniformity correction of infrared images. Infrared detectors are highly susceptible to ambient temperature fluctuations, requiring continuous periodic updates of the correction coefficients.

[0003] The existing non-uniformity correction algorithm is Y = KX + B, where X represents the image data before non-uniformity correction, K represents the gain coefficient, B represents the bias coefficient, and Y represents the image data after non-uniformity correction. The gain coefficient K and bias coefficient B are factory calibration data and can only eliminate image non-uniformity introduced by the detector itself.

[0004] While using shutter image data to replace low-temperature data can mitigate environmental non-uniformity to some extent, for traditional infrared detectors, considering structural design and aesthetics, the shutter can only be located between the lens and the camera mechanism. In this case, simply using the acquired shutter data to replace low-temperature data cannot eliminate the non-uniformity caused by uneven lens and shutter temperatures. Furthermore, in practical applications, ambient temperature changes, leading to uneven shutter plate temperature and affecting the quality of shutter calibration. Even when using a pre-calibrated bias coefficient B during shutter calibration, the temperature changes over time cannot overcome the non-uniformity caused by environmental factors in real-world applications. Summary of the Invention

[0005] This invention addresses the technical problem in existing infrared detection imaging techniques where shutter correction improves image uniformity, but various environmental factors and lens characteristics can cause uneven shutter shutter temperature, introducing additional non-uniform noise. This leads to a significant reduction in image uniformity after shutter correction, affecting image quality. The invention provides an improved adaptive shutter correction method for infrared images.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] An improved adaptive shutter correction method for infrared images, characterized by the following steps:

[0008] Step 1) Calibrate shutter parameters

[0009] 1.1 The infrared detector acquires two blackbody images at high and low temperatures in front of the lens within its operating temperature range, and calculates the gain coefficient K based on the two-point non-uniformity correction method.

[0010] 1.2. Based on the low-temperature blackbody image obtained in step 1.1, calculate the non-uniformity calibration correction parameter Δ(i,j) of the infrared detector:

[0011]

[0012] Where D2(i,j) is the gray value corresponding to the pixel in the i-th row and j-th column of the low-temperature blackbody image;

[0013] Let be the average gray value in the spatial domain corresponding to the pixel in the i-th row and j-th column of the low-temperature blackbody image;

[0014] 1.3 Save the shutter calibration parameters, which include the low-temperature blackbody image data D2(i,j), the gain coefficient K, and the non-uniformity calibration correction parameter Δ(i,j);

[0015] Step 2) Shutter calibration

[0016] 2.1. Load shutter calibration parameters;

[0017] 2.2. Close the shutter, acquire a continuous Num frame image sequence from the image to be corrected, calculate the temporal average image of the image sequence, and use it as the current shutter image data D′(i,j); Nnm>1;

[0018] 2.3 Calculate the correlation between the current shutter image data D′(i,j) and the low temperature blackbody image data D2(i,j), and correct the non-uniformity calibration correction parameter Δ(i,j) according to its correlation to obtain the corrected non-uniformity correction parameter Δ′(i,j);

[0019] 2.4. Traverse the pixels in the image to be corrected, and calculate the corrected image data Y(i,j) by combining the corrected non-uniformity correction parameter Δ′(i,j);

[0020] Y(i,j)=K[X(i,j)-[D′(i,j)-para1×Δ′(i,j)]]+P(i,j)

[0021] Where X(i,j) is the image data of the image to be corrected before correction; para1 is the degree threshold;

[0022] P(i,j) is the mean value of the low-temperature blackbody image data after D2(i,j) correction;

[0023]

[0024] M and N are the number of rows and columns of the blackbody image, respectively, i∈[1,M],j∈[1,N].

[0025] Furthermore, step 2.3 specifically includes:

[0026] 2.3.1 Using the non-uniformity calibration correction parameter Δ(i,j) as template data, calculate the mean Δ_mean(i,j) of the template data:

[0027]

[0028] 2.3.2 Calculate the difference image diff(i,j) between the current shutter speed image data D′(i,j) and the low temperature blackbody image data D2(i,j):

[0029] diff(i,j) = D′(i,j) - D2(i,j)

[0030] 2.3.3 Calculate the mean of the difference image diff(i,j): diff_mean(i,j)

[0031]

[0032] 2.3.4. Divide the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j) into a grid of size winsize_DV×winsize_DH respectively.

[0033] winsize_DV takes the value of a positive integer in the range (1, M) that is divisible by M, and winsize_DH takes the value of a positive integer in the range (1, N) that is divisible by N;

[0034] 2.3.5 Calculate the covariance mix_sum in the non-uniformity calibration correction parameter Δ(i,j) grid and the standard deviation part_mod_sum in the difference image diff(i,j) grid respectively;

[0035]

[0036] Where height_start is the starting row position of the grid, height_end is the last row position of the grid, width_start is the starting column position of the grid, and width_end is the last column position of the grid;

[0037] 2.3.6 Calculate the non-uniformity calibration correction parameter Δ(i,j) and the inter-grid correlation coefficient cor(idx) of the grid corresponding to the location of the difference image diff(i,j):

[0038]

[0039] Where cor(idx) is the inter-grid correlation coefficient of the idx-th grid;

[0040] 2.3.7 Set the correlation threshold Thr and correct the correlation coefficient between grids;

[0041]

[0042] Where, cor′(idx) is the inter-grid correlation coefficient of the idx-th grid after correction;

[0043] 2.3.8 Calculate the global correlation coefficient cor_mean between the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j):

[0044]

[0045] Where k is the number of grid cells.

[0046] count1 is the number of grids whose inter-grid correlation coefficients satisfy cor[idx]≥Thr;

[0047] 2.3.9 Calculate the corrected non-uniformity correction parameter Δ′(i,j):

[0048] Δ′(i,j)=cor_mean×diff(i,j).

[0049] Furthermore, in step 2.2, Num is 32, 64, or 128.

[0050] Furthermore, in step 2.4, para1∈(0,2).

[0051] Furthermore, in step 2.3.4, winsize_DV = winsize_DH = 16.

[0052] Furthermore, in step 2.3.7, Thr∈(0,1).

[0053] Compared with the prior art, the present invention has the following beneficial technical effects:

[0054] The improved adaptive shutter correction method for infrared images provided by this invention uses data from a uniform object collected in front of the lens as the ideal shutter speed data. Building upon existing non-uniformity correction methods, it reconstructs the ideal shutter speed data using shutter speed data from traditional locations, achieving more comprehensive correction of infrared image non-uniformity. This effectively solves the non-uniformity introduced by shutter temperature inhomogeneity and lens factors. Furthermore, the method's shutter speed data correction process is simple, effective, and easy to implement in engineering. Attached Figure Description

[0055] Figure 1 The flowchart illustrates the improved adaptive shutter correction method for infrared images according to the present invention. Detailed Implementation

[0056] To make the objectives, advantages, and features of this invention clearer, the following detailed description of an improved adaptive shutter correction method for infrared images, in conjunction with the accompanying drawings and specific embodiments, is provided. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of this invention and are not intended to limit the scope of protection of this invention.

[0057] like Figure 1 As shown, the present invention provides an improved adaptive shutter correction method for infrared images, comprising two parts: coefficient calibration and shutter correction, specifically including the following steps:

[0058] Step 1) Calibrate shutter parameters

[0059] 1.1 The infrared detector acquires two blackbody images at high and low temperatures in front of the lens within its operating temperature range, and calculates the gain coefficient K based on the two-point non-uniformity correction method.

[0060] Let the high-temperature blackbody image data be D1(i,j) and the low-temperature blackbody image data be D2(i,j), where M and N are the number of rows and columns of the high-temperature or low-temperature blackbody image, respectively, i∈[1,M], j∈[1,N]. According to the two-point non-uniformity correction method y=K×x+B, the high-temperature blackbody image data (x1,y1) and the low-temperature blackbody image data (x2,y2) satisfy:

[0061]

[0062] Solving for the given information yields:

[0063]

[0064] 1.2. Based on the low-temperature blackbody image obtained in step 1.1, calculate the non-uniformity calibration correction parameter Δ(i,j) of the infrared detector:

[0065]

[0066] Where D2(i,j) is the gray value corresponding to the pixel in the i-th row and j-th column of the low-temperature blackbody image;

[0067] Let be the average gray value in the spatial domain corresponding to the pixel in the i-th row and j-th column of the low-temperature blackbody image;

[0068] 1.3 Save the shutter calibration parameters, including the low-temperature blackbody image data D2(i,j), the gain coefficient K, and the non-uniformity calibration correction parameter Δ(i,j);

[0069] Step 2) Shutter calibration

[0070] 2.1. Load shutter calibration parameters;

[0071] 2.2. Close the shutter, acquire a continuous Num frame image sequence from the image to be corrected, calculate the temporal average image of the image sequence, and use it as the current shutter image data D′(i,j); Nnm>1;

[0072] In this embodiment, the temporal average of 128 consecutive frames of the image to be corrected is obtained and used as the current shutter image data D′(i,j);

[0073] 2.3 Calculate the correlation between the current shutter image data D′(i,j) and the low temperature blackbody image data D2(i,j), and correct the non-uniformity calibration correction parameter Δ(i,j) according to its correlation to obtain the corrected non-uniformity correction parameter Δ′(i,j);

[0074] 2.3.1 Using the non-uniformity calibration correction parameter Δ(i,j) as template data, calculate the mean Δ_mean(i,j) of the template data:

[0075]

[0076] 2.3.2 Calculate the difference image diff(i,j) between the current shutter speed image data D′(i,j) and the low temperature blackbody image data D2(i,j):

[0077] diff(i,j) = D′(i,j) - D2(i,j)

[0078] 2.3.3 Calculate the mean of the difference image diff(i,j): diff_mean(i,j)

[0079]

[0080] 2.3.4. Divide the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j) into a grid of size winsize_DV×winsize_DH respectively.

[0081] winsize_DV takes the value of a positive integer in the range (1, M) that is divisible by M, and winsize_DH takes the value of a positive integer in the range (1, N) that is divisible by N; in this embodiment, winsize_DV = winsize_DH = 16;

[0082] 2.3.5 Calculate the covariance mix_sum in the non-uniformity calibration correction parameter Δ(i,j) grid and the standard deviation part_mod_sum in the difference image diff(i,j) grid respectively;

[0083]

[0084] Where height_start is the starting row position of the grid, height_end is the last row position of the grid, width_start is the starting column position of the grid, and width_end is the last column position of the grid;

[0085] 2.3.6 Calculate the non-uniformity calibration correction parameter Δ(i,j) and the inter-grid correlation coefficient cor(idx) of the grid corresponding to the location of the difference image diff(i,j):

[0086]

[0087] Where cor(idx) is the inter-grid correlation coefficient of the idx-th grid;

[0088] 2.3.7 Set the correlation threshold Thr and correct the correlation coefficient between grids;

[0089]

[0090] Where cor′(idx) is the inter-grid correlation coefficient of the idx-th grid after correction; the correlation threshold Thr ∈ (0,1);

[0091] 2.3.8 Calculate the global correlation coefficient cor_mean between the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j):

[0092]

[0093] Where k is the number of grid cells.

[0094] count1 is the number of grids whose inter-grid correlation coefficients satisfy cor[idx]≥Thr;

[0095] 2.3.9 Calculate the corrected non-uniformity correction parameter Δ′(i,j):

[0096] Δ′(i,j)=cor_mean×diff(i,j).

[0097] 2.4. Traverse the pixels in the image to be corrected, and calculate the corrected image data Y(i,j) by combining the corrected non-uniformity correction parameter Δ′(i,j);

[0098] Y(i,j)=K[X(i,j)-[D′(i,j)-para1×Δ′(i,j)]]+P(i,j)

[0099] Where X(i,j) is the image data of the image to be corrected before correction; para1 is the degree threshold, para1∈(0,2);

[0100] P(i,j) is the mean value of the low-temperature blackbody image data after D2(i,j) correction;

[0101]

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.

Claims

1. An improved adaptive shutter correction method for infrared images, characterized by comprising the following steps: Step 1) Calibrate shutter parameters 1.1 The infrared detector acquires two blackbody images at high and low temperatures in front of the lens within its operating temperature range, and calculates the gain coefficient K based on the two-point non-uniformity correction method. 1.

2. Based on the low-temperature blackbody image obtained in step 1.1, calculate the non-uniformity calibration correction parameter Δ(i,j) of the infrared detector: wherein D2(i,j) represents the gray value of the pixel in the i-th row and j-th column of the low-temperature blackbody image. is the spatial average gray value corresponding to the i-th row and j-th column pixel point of the low-temperature black body image. 1.3 Save the shutter calibration parameters, which include the low-temperature blackbody image data D2(i,j), the gain coefficient K, and the non-uniformity calibration correction parameter Δ(i,j); Step 2) Shutter calibration 2.

1. Load shutter calibration parameters; 2.

2. Close the shutter, acquire a continuous Num frame image sequence from the image to be corrected, calculate the temporal average image of the image sequence, and use it as the current shutter image data D′(i,j); Nnm>1; 2.

3. Calculate the correlation between the current shutter speed image data D′(i,j) and the low-temperature blackbody image data D2(i,j), and correct the non-uniformity calibration correction parameter Δ(i,j) based on the correlation to obtain the corrected non-uniformity correction parameter Δ′(i,j); specifically: 2.3.1 Using the non-uniformity calibration correction parameter Δ(i,j) as template data, calculate the mean Δ_mean(i,j) of the template data: 2.3.2 Calculate the difference image diff(i,j) between the current shutter speed image data D′(i,j) and the low temperature blackbody image data D2(i,j): diff(i,j)=D′(i,j)-D2(i,j); 2.3.3 Calculate the mean of the difference image diff(i,j): diff_mean(i,j) 2.3.

4. Divide the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j) into a grid of size winsize_DV×winsize_DH respectively. winsize_DV takes the value of a positive integer in the range (1, M) that is divisible by M, and winsize_DH takes the value of a positive integer in the range (1, N) that is divisible by N; 2.3.5 Calculate the covariance mix_sum in the non-uniformity calibration correction parameter Δ(i,j) grid and the standard deviation part_mod_sum in the difference image diff(i,j) grid respectively; Where height_start is the starting row position of the grid, height_end is the last row position of the grid, width_start is the starting column position of the grid, and width_end is the last column position of the grid; 2.3.6 Calculate the non-uniformity calibration correction parameter Δ(i,j) and the inter-grid correlation coefficient cor(idx) of the grid corresponding to the location of the difference image diff(i,j): Where cor(idx) is the inter-grid correlation coefficient of the idx-th grid; 2.3.7 Set the correlation threshold Thr and correct the correlation coefficient between grids; Where, cor′(idx) is the inter-grid correlation coefficient of the idx-th grid after correction; 2.3.8 Calculate the global correlation coefficient cor_mean between the non-uniformity calibration correction parameter Δ(i,j) and the difference image diff(i,j): where k is the number of grids, count1 is the number of grids whose inter-grid correlation coefficients satisfy cor[idx]≥Thr; 2.3.9 Calculate the corrected non-uniformity correction parameter Δ′(i,j): Δ′(i,j)=cor_mean×diff(i,j); 2.

4. Traverse the pixels in the image to be corrected, and calculate the corrected image data Y(i,j) by combining the corrected non-uniformity correction parameter Δ′(i,j); Y(i,j)=K[X(i,j)-[D′(i,j)-para1×Δ′(i,j)]]+P(i,j); Where X(i,j) is the image data of the image to be corrected before correction; para1 is the degree threshold; P(i,j) is the mean value of the low-temperature blackbody image data after D2(i,j) correction; M and N are the number of rows and columns of the image, respectively, i∈[1,M],j∈[1,N].

2. The method for adaptive shutter correction of improved infrared images according to claim 1, characterized in that: In step 2.2, Num is 32, 64, or 128.

3. The improved adaptive shutter correction method for infrared images according to claim 2, characterized in that: In step 2.4, para1∈(0,2).

4. The improved adaptive shutter correction method for infrared images according to claim 1, characterized in that: In step 2.3.4, winsize_DV = winsize_DH = 16.

5. The improved adaptive shutter correction method for infrared images according to claim 1, characterized in that: In step 2.3.7, Thr∈(0,1).

Images