A detector non-uniformity correction parameter setting and detection method

Abstract

A method for setting and detecting detector non-uniformity correction parameters relates to the technical field of detector non-uniformity correction and solves the problem that, during existing detector non-uniformity correction, deviations between a correction coefficient and an input image cannot be detected. The correction system of the present invention comprises a TDICMOS detector, an imaging controller, an integrating sphere, an image data acquisition and processor, and a temperature controller. The detection method sets the bit widths of the decimal part of the product correction coefficient and the addition and subtraction correction coefficients according to the bit width of the input parallel image data, and sets the bit width of the integer part of the single product correction coefficient according to the multiple of the difference between the maximum and minimum values of the acquired image. During the verification and detection process, a deviation of 1 DN value is tolerated, but coefficient values whose calculation results exceed 1 DN are set in adjacent correction coefficients.

Description

Technical Field

[0001] The present invention relates to a method for setting and detecting non-uniformity correction parameters of a detector, and in particular to a method for setting and detecting non-uniformity correction parameters of a detector based on space applications. Background Art

[0002] When performing detector non-uniformity correction, the more decimal places you set, the higher the accuracy; the more integer places you set, the wider the correction range. However, the more bits you set, the more computing resources and storage resources you use for non-uniformity correction. Therefore, a trade-off between computational bit width and resource utilization is necessary. The result after non-uniformity correction may be a decimal. Due to differences in bit width across different hardware platforms, the final calculation result may differ by as much as one DN. If the difference of this 1 DN value is not tolerated, a large number of errors may occur only due to the difference in the clipping bit width; if the error is tolerated, the difference in the final value between adjacent pixels does not exceed 1 DN value. Even if the relative position of the correction coefficient and the input image deviates by one position during the correction process (for example, the correction coefficient and the image should correspond one to one. If a row has 1024 pixels, the address corresponding to the image is 0-1023, and the address corresponding to the correction coefficient should also be 0-1023; the image and correction coefficient should correspond to each other (i is 0-1023). If an operation error occurs, the image at the i-th address may correspond to the correction coefficient at the i+1 or i-1 address), it is difficult to detect.

[0003] To this end, the present invention proposes a detector non-uniformity correction setting and detection method. The bit widths of the decimal portion of the multiplication correction coefficient and the addition and subtraction correction coefficients are set based on the bit width of the input parallel image data, and the bit width of the integer portion of the single-product correction coefficient is set based on the multiple of the difference between the maximum and minimum values of the acquired image. During the verification and detection process, a deviation of one DN value is tolerated, but coefficient values exceeding one DN are set between adjacent correction coefficients. Summary of the Invention

[0004] The present invention aims to solve the problem that when performing the existing detector non-uniformity correction, when there is a deviation between the correction coefficient and the input image, it cannot be detected, and provides a detector non-uniformity correction setting and detection method.

[0005] A detector non-uniformity correction setting and detection method is implemented using a TDICMOS data correction system. The method sets the decimal part of the product correction coefficient and the bit width of the addition and subtraction correction coefficient c according to the bit width n of the parallel image data of the TDICMOS detector. The product correction coefficient consists of a quadratic correction coefficient a and a primary correction coefficient b.

[0006] The correction formula of the output image after correction by the imaging controller is:

[0007] y=ax 2 +bx+c

[0008] Where x is the DN value of the input image, and y is the DN value of the corrected image;

[0009] The decimal part of the quadratic correction coefficient a occupies 2n binary digits, the positive and negative signs occupy 1 bit, and the integer part occupies 1 bit; the positive and negative signs of the addition and subtraction correction coefficient c occupy 1 bit, and the integer part occupies n binary digits; the decimal part of the primary correction coefficient b occupies n+1 binary digits, and the integer part occupies k binary digits, that is, the decimal number represented by the integer part is required to be greater than the maximum image DN value DN obtained by acquisition max and the minimum image DN value DN min The ratio of

[0010] The correctness detection process of the non-uniformity correction process is as follows:

[0011] Step 1: Set the DN value x of the input image to a constant 2 n-1 , the secondary correction coefficient a and the primary correction coefficient b are both 0, and the addition and subtraction correction coefficient c ranges from -2 n to 2 n Change 1 bit by bit to perform cyclic changes; check whether the DN value y of the corrected image is correct;

[0012] Step 2: Set the DN value x of the input image to a constant 2 n-1 , the addition and subtraction correction coefficient c and the primary correction coefficient b are both 0, and the value of the secondary correction coefficient a is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct;

[0013] Step 3: Set the DN value x of the input image to a constant 2 n-k-1 , the addition and subtraction correction coefficient c and the secondary correction coefficient a are both 0, and the value of the primary correction coefficient b is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct;

[0014] Step 4: Set the DN value x of the input image to be from 2 n -1 to 0 and changes cyclically, bit by bit changing by 1; the value of the secondary correction coefficient a is from arrive Perform cyclic changes, bit by bit The value of the primary correction coefficient b is arrive Perform cyclic changes, bit by bit The value of c ranges from -2 n to 2 n Change 1 bit by bit in a cyclic manner; check whether the DN value y of the corrected image is correct.

[0015] Beneficial effects of the present invention:

[0016] 1. The bit widths of the decimal part of the multiplication correction coefficient and the addition and subtraction correction coefficients are set based on the bit width of the input parallel image data. This minimizes resource usage while ensuring maximum correction accuracy. 2. The bit width of the integer part of the single multiplication correction coefficient is set based on the multiple of the difference between the maximum and minimum values of the acquired image. This minimizes resource usage while ensuring a sufficient correction range.

[0017] 3. During the verification process, a deviation of 1 DN value is tolerated, but the coefficient value of the calculated result exceeding 1DN is set in the adjacent correction coefficients. This can avoid misjudgment caused by the difference in the interception bit width and detect the deviation between the correction coefficient and the input image position. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a principle block diagram of a TDICMOS data correction system in a detector non-uniformity correction parameter setting and detection method described in the present invention. DETAILED DESCRIPTION

[0019] Combine Figure 1 This embodiment describes a detector non-uniformity correction setup and detection method implemented using a TDICMOS data correction system. The system includes a TDICMOS detector, an imaging controller, an integrating sphere, an image data acquisition and processor, and a temperature controller. The imaging controller generates the control and drive signals required for the TDICMOS detector to operate, and the TDICMOS detector returns relevant status signals. The serial image data output by the TDICMOS detector is processed by the imaging controller and then fed as parallel image data to the image data acquisition and processor for calculation of imaging and correction parameters. The TDICMOS detector and imaging controller are placed within a temperature controller to achieve constant temperature control.

[0020] In this embodiment, the bit widths of the decimal part of the product correction coefficients (a and b) and the addition and subtraction correction coefficients c are set according to the bit width n of the detector parallel image data.

[0021] The correction formula of the output image after correction by the imaging controller is:

[0022] y=ax 2+bx+c

[0023] Where x is the DN value of the input image, y is the DN value of the corrected image, a is the quadratic correction coefficient, b is the primary correction coefficient, and c is the addition and subtraction correction coefficient.

[0024] The decimal part of the quadratic correction coefficient a occupies 2n binary digits, the positive and negative signs occupy 1 bit, and the integer part occupies 1 bit; the positive and negative signs of the addition and subtraction correction coefficient c occupy 1 bit, and the integer part occupies n binary digits; the decimal part of the primary correction coefficient b occupies n+1 binary digits, and the integer part occupies k binary digits, that is, the decimal number represented by the integer part is required to be greater than the maximum image DN value DN obtained by acquisition max and the minimum image DN value DN min That is:

[0025]

[0026] In this embodiment, the steps for checking the correctness of the non-uniformity correction process are as follows:

[0027] Step 1: The DN value x of the input image is constant 2 n-1 , the correction coefficients a and b are 0, and the value of c ranges from -2 n to 2 n Change 1 bit by bit in a cyclic manner; check whether the DN value y of the corrected image is correct.

[0028] Step 2: The DN value x of the input image is constant 2 n-1 , the correction coefficients c and b are 0, and the value of a is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct.

[0029] Step 3: The DN value x of the input image is constant 2 n-k-1 , the correction coefficients c and a are 0, and the value of b is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct.

[0030] Step 4: The DN value x of the input image is from 2 n -1 to 0 and changes cyclically, bit by bit changing by 1; the value of the correction coefficient a is from arrive Perform cyclic changes, bit by bit The value of b is arrive Perform cyclic changes, bit by bit The value of c ranges from -2n to 2 n Change 1 bit by bit in a cyclic manner; check whether the DN value y of the corrected image is correct.

[0031] In this embodiment, the imaging controller uses a Virtex 6 device and its internal resources; the TDICMOS image sensor uses a customized product from Gpixel; the integrating sphere uses a common integrating sphere with an aperture larger than the detector size; the image data acquisition and processor uses a PC with a CameraLink acquisition card for image acquisition and processing; and the temperature controller uses a high and low temperature chamber.

Claims

1. A method for setting and detecting detector non-uniformity correction parameters, implemented by a TDICMOS data correction system, characterized by: The decimal part of the product correction coefficient and the bit width of the addition and subtraction correction coefficient c are set according to the bit width n of the parallel image data of the TDICMOS detector, wherein the product correction coefficient consists of a quadratic correction coefficient a and a primary correction coefficient b; The correction formula of the output image after correction by the imaging controller is: y=ax 2 +bx+c Where x is the DN value of the input image, and y is the DN value of the corrected image; The decimal part of the quadratic correction coefficient a occupies 2n binary digits, the positive and negative signs occupy 1 bit, and the integer part occupies 1 bit; the positive and negative signs of the addition and subtraction correction coefficient c occupy 1 bit, and the integer part occupies n binary digits; the decimal part of the primary correction coefficient b occupies n+1 binary digits, and the integer part occupies k binary digits, that is, the decimal number represented by the integer part is required to be greater than the maximum image DN value DN obtained by acquisition max and the minimum image DN value DN min The ratio of The correctness detection process of the non-uniformity correction process is as follows: Step 1: Set the DN value x of the input image to a constant 2 n-1 , the secondary correction coefficient a and the primary correction coefficient b are both 0, and the addition and subtraction correction coefficient c ranges from -2 n to 2 n Change 1 bit by bit to perform cyclic changes; check whether the DN value y of the corrected image is correct; Step 2: Set the DN value x of the input image to a constant 2 n-1 , the addition and subtraction correction coefficient c and the primary correction coefficient b are both 0, and the value of the secondary correction coefficient a is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct; Step 3: Set the DN value x of the input image to a constant 2 n-k-1 , the addition and subtraction correction coefficient c and the secondary correction coefficient a are both 0, and the value of the primary correction coefficient b is arrive Perform cyclic changes, bit by bit Check whether the DN value y of the corrected image is correct; Step 4: Set the DN value x of the input image to be from 2 n -1 to 0 and changes cyclically, bit by bit changing by 1; the value of the secondary correction coefficient a is from arrive Perform cyclic changes, bit by bit The value of the primary correction coefficient b is arrive Perform cyclic changes, bit by bit The value of c ranges from -2 n to 2 n Change 1 bit by bit in a cyclic manner; check whether the DN value y of the corrected image is correct.

2. The method for setting and detecting detector non-uniformity correction parameters according to claim 1, characterized in that: The TDICMOS data correction system includes a TDICMOS detector, an imaging controller, an integrating sphere, and an image data acquisition and processor. The imaging controller generates the control and drive signals required for the operation of the TDICMOS detector, and the TDICMOS detector returns relevant status signals. The serial image data output by the TDICMOS detector is processed by the imaging controller and then sent to the image data acquisition and processor in the form of parallel image data for processing, thereby realizing the calculation of imaging parameters and correction parameters.

3. The method for setting and detecting detector non-uniformity correction parameters according to claim 1, wherein: A temperature controller is also included. The TDICMOS detector and the imaging controller are placed in the temperature controller to achieve constant temperature control.

Images