A method for detecting and compensating the position of light spot fault points based on register judgment

Abstract

The present invention relates to a method for detecting and compensating the position of a light spot fault point based on register judgment. When calculating the centroid of an image, the detection result of the light spot fault point is read. A light spot fault point judgment step of a pixel point is added after the threshold judgment step in the centroid calculation process. The pixel points whose amplitude f is greater than 0 and are not light spot fault points are detected. The coordinates of these pixel points are arranged in sequence as (x1, y1), (x2, y2), (x3, y3), ..., (x i ,y i ), the horizontal coordinate of the centroid of the image is: #imgabs0# and the vertical coordinate is: #imgabs1# When calculating the image weight, read the detection result of the light spot fault point, and add a light spot fault point judgment step of a pixel point in the above center of gravity calculation process. When the pixel point is the light spot fault point, record the amplitude f(a,b) as 0, otherwise record its amplitude f(a,b) unchanged, and then complete the coordinate calculation of the center of gravity according to the above formula.

Description

Technical field:

[0001] The present invention relates to the field of image processing technology, and in particular to a light spot fault point position detection and compensation method based on register judgment, which can eliminate the influence of light spots on the calculation of image centroid and center of gravity. Background technology:

[0002] The method for calculating the centroid and center of gravity in image processing is as follows: As shown in Equation 1, assuming that the image is composed of an m×n pixel matrix, it can be defined as a two-dimensional function f(x, y), where (x, y) are the coordinates of each pixel (x = 1, 2, 3...m; y = 1, 2, 3...n). The amplitude f at any coordinate (x, y) is defined as the brightness of the image at that location and also represents the mass of the object. The f value of each pixel is an eight-bit binary number ranging from 0 to 255. The larger the pixel f value, the greater the mass of the object and the brighter the image.

[0003] Formula 1, calculation method of centroid:

[0004] The amplitude f of each pixel in the image is thresholded. If f is greater than 0, the coordinate value of the pixel is recorded. If f is less than 0, no processing is done. Assume that there are a total of i pixels with amplitude f greater than 0. The coordinates of these pixels are arranged in order as (x1, y1), (x2, y2), (x3, y3), ..., (x i ,y i ), then the abscissa of the centroid of the image is: (j=1,2,3,……,i), the vertical coordinate is: (j=1,2,3,……,i).

[0005] The calculation method of the center of gravity is as follows: Since the amplitude f of each pixel represents both the brightness and the mass of the point, the center of gravity of the image can be calculated by calculating the amplitude f. The horizontal coordinate of the center of gravity is (a=1,2,3,……,m), (b=1,2,3,……,n); the vertical coordinate of the center of gravity is (a=1,2,3,...,m), (b=1,2,3,...,n).

[0006] In practical applications, prolonged high-energy particle impacts on sensors can cause sensor failure. Pixels in the absence of objects have an amplitude of f = 0. However, when a sensor has a spot failure, the pixel becomes brighter in the absence of the object, and the amplitude of the pixel increases, even reaching the saturation value of 255. This affects the calculation of the image's centroid and center of gravity, causing the calculated horizontal and vertical coordinates to shift toward the spot of the spot failure. Summary of the invention:

[0007] In view of the problems existing in the prior art, the present invention proposes a light spot fault point position detection and compensation method based on register judgment, which can effectively correct image processing deviation.

[0008] The present invention is achieved by the following measures:

[0009] A method for detecting and compensating a light spot fault point position based on register judgment, characterized by comprising the following steps:

[0010] Step 1: Construct a detection module, which consists of three FIFOs with a storage depth of m-3 and a bit width of 8 bits and nine 8-bit registers. Assume that the image resolution is m×n, and the image data is input into FIFO1, register 1, register 2, register 3, FIFO2, register 4, register 5, register 6, FIFO3, register 7, register 8, and register 9 in sequence;

[0011] Step 2: Determine whether the current pixel is a fault point: write data to the detection module in sequence. When registers 1-9 are written with f(1, 1), f(2, 1), f(3, 1), f(1, 2), f(2, 2), f(3, 2), f(1, 3), f(2, 3), and f(3, 3), respectively, calculate the difference between the values stored in register 5 and the other eight registers. When all eight differences are greater than the threshold, the pixel f(2, 2) stored in register 5 is judged to be a light spot fault point. Otherwise, the pixel is judged to be a normal point, and the threshold is adjustable.

[0012] Step 3: Calculate and judge each subsequent pixel point until registers 1-9 are written with f(m-2, n-2), f(m-1, n-2), f(m, n-2), f(m-2, n-1), f(m-1, n-1), f(m, n-1), f(m-2, n), f(m-1, n), and f(m, n). The last calculation is completed to determine whether the pixel point f(m-1, n-1) is a light spot fault point. After storing the coordinates of the pixel point determined to be a light spot fault point, wait for the next set of image data for calculation and judgment.

[0013] Step 4: Compensation for image spot fault points: Compensate and calculate the centroid and center of gravity of the image with a resolution of m×n:

[0014] Step 4-1: When calculating the image centroid, read the detection results of the light spot fault point. Add a light spot fault point judgment step of one pixel after the threshold judgment step in the centroid calculation process. Detect the pixels whose amplitude f is greater than 0 and are not light spot fault points. Arrange the coordinates of these pixels in sequence as (x1, y1), (x2, y2), (x3, y3), ..., (x i ,y i ), then the abscissa of the centroid of the image is: (j=1,2,3,……,i), the vertical coordinate is: (j=1,2,3,……,i);

[0015] Step 4-2: The compensation calculation process of the image center of gravity is as follows: When calculating the image weight, read the detection result of the detection spot fault point, and add a pixel point spot fault point judgment step in the above center of gravity calculation process. The original calculation formula is: horizontal coordinate (a=1,2,3,……,m), (b=1,2,3,……,n); vertical coordinate (a=1,2,3,……,m), (b=1,2,3,……,n), add a compensation comparator. When the pixel point is a light spot fault point, record the amplitude f(a,b) as 0, otherwise record its amplitude f(a,b) unchanged, and then complete the coordinate calculation of the center of gravity according to the above formula.

[0016] The present invention can effectively and automatically monitor and eliminate light spot fault points. The existing technical means completely use trigger resources and require a large FPGA to complete. The method of the present invention uses internal RAM resources to build a FIFO, which can effectively reduce the use of FPGA internal triggers, greatly reduce the cost of use, and can be achieved with a small-capacity FPGA. Description of the drawings:

[0017] Attachment Figure 1 It is a structural diagram of the detection module in the present invention. Specific implementation method:

[0018] The present invention will be further described below with reference to the accompanying drawings.

[0019] The present invention proposes a method for detecting and compensating the position of a light spot fault point based on register judgment, which is characterized by comprising the following steps:

[0020] Step 1: Build the detection module, as shown in the attached Figure 1As shown in the figure, the detection module consists of three FIFOs with a storage depth of m-3 and a bit width of 8 bits and nine 8-bit registers. Assuming the image resolution is m×n, the image data is input into FIFO1, register 1, register 2, register 3, FIFO2, register 4, register 5, register 6, FIFO3, register 7, register 8, and register 9 in sequence;

[0021] Step 2: Determine whether the current pixel is a fault point: write data to the detection module in sequence. When registers 1-9 are written with f(1, 1), f(2, 1), f(3, 1), f(1, 2), f(2, 2), f(3, 2), f(1, 3), f(2, 3), and f(3, 3), respectively, calculate the difference between the values stored in register 5 and the other eight registers. When all eight differences are greater than the threshold, the pixel f(2, 2) stored in register 5 is judged to be a light spot fault point. Otherwise, the pixel is judged to be a normal point, and the threshold is adjustable.

[0022] Step 3: Calculate and judge each subsequent pixel point until registers 1-9 are written with f(m-2, n-2), f(m-1, n-2), f(m, n-2), f(m-2, n-1), f(m-1, n-1), f(m, n-1), f(m-2, n), f(m-1, n), and f(m, n). The last calculation is completed to determine whether the pixel point f(m-1, n-1) is a light spot fault point. After storing the coordinates of the pixel point determined to be a light spot fault point, wait for the next set of image data for calculation and judgment.

[0023] Step 4: Compensation for image spot fault points: Compensate and calculate the centroid and center of gravity of the image with a resolution of m×n:

[0024] Step 4-1: When calculating the image centroid, read the detection results of the light spot fault point. Add a light spot fault point judgment step of one pixel after the threshold judgment step in the centroid calculation process. Detect the pixels whose amplitude f is greater than 0 and are not light spot fault points. Arrange the coordinates of these pixels in sequence as (x1, y1), (x2, y2), (x3, y3), ..., (x i ,y i ), then the abscissa of the centroid of the image is: (j=1,2,3,……,i), the vertical coordinate is: (j=1,2,3,……,i);

[0025] Step 4-2: The compensation calculation process of the image center of gravity is as follows: When calculating the image weight, read the detection result of the detection spot fault point, and add a pixel point spot fault point judgment step in the above center of gravity calculation process. The original calculation formula is: horizontal coordinate (a=1,2,3,……,m), (b=1,2,3,……,n); vertical coordinate (a=1,2,3,……,m), (b=1,2,3,……,n), add a compensation comparator. When the pixel point is a light spot fault point, record the amplitude f(a,b) as 0, otherwise record its amplitude f(a,b) unchanged, and then complete the coordinate calculation of the center of gravity according to the above formula.

[0026] The present invention can effectively and automatically monitor and eliminate light spot fault points. The existing technical means completely use trigger resources and require a large FPGA to complete. The method of the present invention uses internal RAM resources to build a FIFO, which can effectively reduce the use of FPGA internal triggers, greatly reduce the cost of use, and can be achieved with a small-capacity FPGA.

Claims

1. A method for detecting and compensating the position of a light spot fault point based on register judgment, characterized in that: The following steps are involved: Step 1: Construct a detection module, which consists of three FIFOs with a storage depth of m-3 and a bit width of 8 bits and nine 8-bit registers. Assume that the image resolution is m×n, and the image data is input into FIFO1, register 1, register 2, register 3, FIFO2, register 4, register 5, register 6, FIFO3, register 7, register 8, and register 9 in sequence; Step 2: Determine whether the current pixel is a fault point: write data to the detection module in sequence. When registers 1-9 are written with f(1, 1), f(2, 1), f(3, 1), f(1, 2), f(2, 2), f(3, 2), f(1, 3), f(2, 3), and f(3, 3), respectively, calculate the difference between the values stored in register 5 and the other eight registers. When all eight differences are greater than the threshold, the pixel f(2, 2) stored in register 5 is judged to be a light spot fault point. Otherwise, the pixel is judged to be a normal point, and the threshold is adjustable. Step 3: Calculate and judge each subsequent pixel point until registers 1-9 are written with f(m-2, n-2), f(m-1, n-2), f(m, n-2), f(m-2, n-1), f(m-1, n-1), f(m, n-1), f(m-2, n), f(m-1, n), and f(m, n). The last calculation is completed to determine whether the pixel point f(m-1, n-1) is a light spot fault point. After storing the coordinates of the pixel point determined to be a light spot fault point, wait for the next set of image data for calculation and judgment. Step 4: Compensation for image spot fault points: Compensate and calculate the centroid and center of gravity of the image with a resolution of m×n: Step 4-1: When calculating the image centroid, read the detection results of the light spot fault point. Add a light spot fault point judgment step of one pixel after the threshold judgment step in the centroid calculation process. Detect the pixels whose amplitude f is greater than 0 and are not light spot fault points. Arrange the coordinates of these pixels in sequence as (x1, y1), (x2, y2), (x3, y3), ..., (x i ,y i ), then the abscissa of the centroid of the image is: The vertical axis is: Step 4-2: The compensation calculation process of the image center of gravity is as follows: When calculating the image center of gravity, read the detection result of the detection spot fault point, and add a pixel point spot fault point judgment step in the above center of gravity calculation process. The original calculation formula is: horizontal coordinate vertical axis Add a compensation comparator. When the pixel point is a light spot fault point, record the amplitude f(a, b) as 0. Otherwise, record the amplitude f(a, b) unchanged. Then, calculate the coordinates of the center of gravity according to the above formula.

Images