Method for detecting abnormal pixels of detector

Abstract

The invention discloses a method for detecting abnormal pixels of a detector, and the method comprises the following steps: selecting a completely shading experiment environment, and guaranteeing that a lens of the detector is completely covered, so as to avoid the interference of stray light; the detector is started for standby, so that the working state of the detector is stable; acquiring dark field images of the detector under the minimum, middle and maximum exposure time; performing abnormal pixel detection on each pixel in the image, and determining whether the current detected pixel is an abnormal pixel or not by adopting a method of combining feature judgment with threshold judgment; the detected abnormal pixels can be marked for subsequent image processing or analysis. By means of the method, the abnormal pixels in the detector can be effectively detected and recognized, the number change of the abnormal pixels of the detector can be monitored regularly, the image quality and reliability are guaranteed, and the method is suitable for CCD, EMCCD and CMOS detectors and has wide application prospects.

Description

Technical Field

[0001] The invention relates to the technical field of photoelectric component testing, and in particular to a method for detecting abnormal pixels of a detector. Background Art

[0002] Abnormal pixels refer to pixels that appear in images of detectors such as CCD, EMCCD and CMOS, and have values ​​significantly different from those of the surrounding normal pixels. This phenomenon may be caused by dark current in the detector, damage to the photosensitive element or other factors. At present, the detection technologies for abnormal pixels in detectors mainly include dynamic laser scanning analysis (DALS), electro-optical detection / electro-optical frequency imaging and other methods. These methods use the principle of electro-optical effect (EO) to analyze and evaluate the failure mechanism of abnormal pixel detection of the chip through the interaction between the laser beam and the surface or internal structure of the chip. However, these technologies usually require professional equipment, have long cycles and high costs, and cannot simulate all complex conditions in actual use scenarios, such as different lighting environments and temperature changes. In addition, long-term or frequent irradiation experiments may cause damage to the focal plane of the detector, thereby reducing efficiency. Summary of the invention

[0003] The object of the present invention is to provide a method for detecting abnormal pixels of a detector. The method has high inspection accuracy and can effectively detect abnormal pixels of detectors such as CCD, EMCCD and CMOS.

[0004] The technical solution adopted by the present invention to solve its technical problem is:

[0005] A method for detecting abnormal pixels of a detector comprises the following steps:

[0006] Step 1: Start the detector in a completely light-shielded environment and put it into standby mode to stabilize the working state of the detector;

[0007] Step 2: Obtain dark field images of the detector at minimum exposure time, medium exposure time, and maximum exposure time;

[0008] Step 3: Detect the change of the DN value of the pixel in the dark field image in the order of minimum exposure time, medium exposure time, and maximum exposure time, and calculate and determine whether the current pixel being tested is an abnormal pixel based on this.

[0009] Furthermore, the method further includes step 4, namely marking the detected abnormal pixels for subsequent image processing or analysis.

[0010] Furthermore, step 1 includes placing the detector to be tested in a darkroom environment, covering the lens of the detector to prevent interference from stray light, powering on the detector, and cooling the detector to the operating temperature so that the detector reaches a stable working state; the darkroom environment and the stable working state mean that the detector is tested without interference from stray light, the ambient temperature remains stable, and the detector is in a preset cooling temperature state.

[0011] Furthermore, the step 2 includes determining the maximum exposure time expt of the detector under test in the linear response range. max , medium exposure time expt mid And the minimum exposure time expt min , expt mid The size is 0.5*(expt min +expt max ), according to expt min , expt mid , expt max The dark field images are collected in the order of exposure time, and no less than n dark field images are collected at each exposure time.

[0012] Furthermore, the number of dark field images in step 2 satisfies n≥20.

[0013] Furthermore, the step 3 comprises the following steps:

[0014] Step 3-1: Average the dark field images at each exposure time to obtain the dark field mean image Dark_master at different exposure times expt (i, j), and according to expt min , expt mid , expt max The order of each Dark_master expt (i, j) is made into a cube file, where (i, j) represents the pixel in the i-th row and j-th column of the image;

[0015] Step 3-2: Detect the DN value of the pixel at position (i, j) in the cube file according to expt min , expt mid , expt max According to the change of the order of , determine whether the current pixel is an abnormal pixel;

[0016] Step 3-3: Traverse all pixels in the cube file according to the method in step 3-2, and obtain the position and total number of abnormal pixels in the image.

[0017] Furthermore, the cube file in step 3-1 is Dark_masterexpt (i, j) according to expt min , expt mid , expt max The order is stacked into a three-dimensional array.

[0018] Furthermore, the determination method in step 3-2 includes:

[0019] Step 3-2-1: Determine whether the DN values ​​of the three pixels at the same position of the cube are monotonically increasing;

[0020] Step 3-2-2: If it is not monotonically increasing, it is judged as an abnormal pixel and the position of the point is marked;

[0021] Step 3-2-3: If it is monotonically increasing, solve the slope of the three pixels at the same position according to the relationship y=ax+b, and average the slopes of all N monotonically increasing points, and then calculate the root mean square error σ. If the difference between the slope of the pixel point at a certain position and the slope mean is greater than 3σ, it is defined as an abnormal pixel point; the method for calculating the root mean square error σ refers to the following formula:

[0022] .

[0023] The beneficial effects of the present invention are:

[0024] The method of the present invention can effectively detect and identify abnormal pixels in the detector, help to regularly monitor the change in the number of abnormal pixels in the detector, and ensure image quality and reliability. By regularly detecting abnormal pixels in the detector, the detector usage scenarios can be reasonably arranged to extend the service life of the detector. At the same time, the image quality can be improved to ensure the reliability and accuracy of the image. The method of the present invention is applicable to CCD, EMCCD and CMOS detectors and has a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flow chart of the method of the present invention;

[0026] Figure 2 1 is an actual detection effect diagram of an embodiment of the present invention; (a) is the original image, and (b) is the detection effect diagram. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0028] like Figure 1 As shown, the present invention provides a method for detecting abnormal pixels of a detector, comprising the following steps:

[0029] Step (1) Place the detector to be tested in a darkroom environment, cover the lens of the detector to prevent interference from stray light, power on the detector, and cool the detector to the operating temperature so that the detector reaches a stable working state; the darkroom environment and stable working state mean that the detector is tested without interference from stray light, the ambient temperature remains stable, and the detector is at a preset cooling temperature. In this step, stray light should be strictly controlled to ensure the effect of subsequent calibration.

[0030] Step (2): determine the maximum exposure time expt of the detector under test in the linear response range. max , medium exposure time expt mid And the minimum exposure time expt min , expt mid The size is 0.5*(expt min +expt max ), according to expt min , expt mid , expt max The dark field images are collected in the order of exposure time, and no less than n dark field images (n ≥ 20) are collected at each exposure time; the exposure time expt min Represents the shortest exposure time of the detector to be tested, expt max Represents the maximum exposure time of the detector under test working in the linear region, expt mid Represents between expt min and expt max This step should ensure that the maximum exposure time expt max Under these conditions, the detector is still in the linear response range and the collected image is not saturated or overflowed.

[0031] Step (3), calculate the average of the dark field images at each exposure time to obtain the dark field average image at different exposure times (Dark_master expt (i, j)), and follow expt min , expt mid , expt max The order of each Dark_master expt (i, j) is made into a cube file; (i, j) represents the pixel in the i-th row and j-th column of the image; the cube file is Dark_master expt (i, j) according to expt min , expt mid , expt max The method of calculating the dark field average image of n dark field images under the same exposure condition in this step refers to the following formula:

[0032]

[0033] Step (4), detect the DN value of the pixel at position (i, j) in the cube file according to expt min , expt mid , expt max The order of changes is determined, and based on this, it is calculated to determine whether the current pixel being tested is an abnormal pixel. The specific judgment method is: ① Determine whether the DN values ​​of the three pixels at the same position of the cube are monotonically increasing; ② If it is not monotonically increasing, it is judged as an abnormal pixel and the position of the point is marked; ③ If it is monotonically increasing, the slope of the three pixels at the same position is solved according to the relationship y=ax+b, and the slope of all monotonically increasing points (assuming there are N) is averaged, and then the root mean square error σ is calculated. If the difference between the slope of the pixel point at a certain position and the mean of the slope is greater than 3σ, it is defined as an abnormal pixel point. The method for calculating the root mean square error σ refers to the following formula:

[0034]

[0035] Step (5): traverse all pixels in the cube file according to the method in step (4), and mark the position and total number of abnormal pixels in the image.

[0036] Figure 2 This is the actual detection effect diagram according to the above method. Compared with the original image, the total number of abnormal pixels detected is 381.

[0037] In summary, the present invention provides a method for detecting abnormal pixels of a detector, comprising the following steps: a) selecting a completely light-shielded experimental environment and ensuring that the detector lens is completely covered to avoid interference from stray light; b) starting the CCD detector to standby to stabilize the working state of the detector; c) obtaining the CCD detector expt min , expt mid , expt maxDark field image under exposure time; d) abnormal pixel detection is performed on each pixel in the image, and a method of feature discrimination combined with threshold judgment is used to determine whether the current pixel being tested is an abnormal pixel; e) the detected abnormal pixels can be marked for subsequent image processing or analysis. Through this method, abnormal pixels in CCD detectors can be effectively detected and identified, which helps to regularly monitor the changes in the number of abnormal pixels in the detector and ensure image quality and reliability. The method is applicable to CCD, EMCCD and CMOS detectors and has broad application prospects. Compared with the prior art, the method of the present invention identifies abnormal pixels in the detector by taking dark field images and performing pixel numerical analysis. During the inspection process, these abnormal pixels can be selected to be filtered or marked to ensure the accuracy and clarity of the image quality. Regularly monitoring the status of the detector chip, reasonably arranging its use scenarios, and extending its service life are of great significance to improving the overall performance. By timely discovering and processing abnormal pixels, the image quality can be significantly improved, thereby ensuring image reliability and accuracy, which is particularly critical in applications with high requirements for image quality. This method not only improves work efficiency, but also can effectively detect images acquired under actual working conditions without affecting the life of the detector itself.

[0038] The above is only a preferred embodiment of the present invention, and does not limit the present invention in any form; any technician familiar with the art can make many possible changes and modifications to the technical solution of the present invention by using the above disclosed methods and technical contents without departing from the scope of the technical solution of the present invention, or modify it into an equivalent embodiment of equivalent changes. Therefore, any simple modification, equivalent replacement, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of protection of the technical solution of the present invention.

Claims

1. A method for detecting abnormal pixels of a detector, characterized in that: The following steps are involved: Step 1: Start the detector in a completely light-shielded environment and put it into standby mode to stabilize the working state of the detector; Step 2: Obtain dark field images of the detector at minimum exposure time, medium exposure time, and maximum exposure time; Step 3: Detect the change of the DN value of the pixel in the dark field image in the order of minimum exposure time, medium exposure time, and maximum exposure time, and calculate and determine whether the current pixel being tested is an abnormal pixel based on this.

2. The method for detecting abnormal pixels of a detector according to claim 1, characterized in that: The method further includes step 4, which is marking the detected abnormal pixels for subsequent image processing or analysis.

3. The method for detecting abnormal pixels of a detector according to claim 1, characterized in that: The step 1 includes placing the detector to be tested in a darkroom environment, covering the lens of the detector to prevent interference from stray light, powering on the detector, and cooling the detector to an operating temperature so that the detector reaches a stable working state; the darkroom environment and the stable working state mean that the detector is tested without interference from stray light, the ambient temperature remains stable, and the detector is in a preset cooling temperature state.

4. The method for detecting abnormal pixels of a detector according to claim 1, characterized in that: The step 2 includes determining the maximum exposure time expt of the detector under test in the linear response range. max , medium exposure time expt mid And the minimum exposure time expt min , expt mid The size is 0.5*(expt min +expt max ), according to expt min , expt mid , expt max The dark field images are collected in the order of exposure time, and no less than n dark field images are collected at each exposure time.

5. The method for detecting abnormal pixels of a detector according to claim 4, characterized in that: The number of dark field images in step 2 satisfies n≥20.

6. The method for detecting abnormal pixels of a detector according to claim 1, characterized in that: The step 3 comprises the following steps: Step 3-1: Average the dark field images at each exposure time to obtain the dark field mean image Dark_master at different exposure times expt (i, j), and according to expt min , expt mid , expt max The order of each Dark_master expt (i, j) is made into a cube file, where (i, j) represents the pixel in the i-th row and j-th column of the image; Step 3-2: Detect the DN value of the pixel at position (i, j) in the cube file according to expt min , expt mid , expt max According to the change of the order of , determine whether the current pixel is an abnormal pixel; Step 3-3: Traverse all pixels in the cube file according to the method in step 3-2, and obtain the position and total number of abnormal pixels in the image.

7. The method for detecting abnormal pixels of a detector according to claim 6, characterized in that: The cube file in step 3-1 is Dark_master expt (i, j) according to expt min , expt mid , expt max The order is stacked into a three-dimensional array.

8. The method for detecting abnormal pixels of a detector according to claim 6, characterized in that: The determination method in step 3-2 includes: Step 3-2-1: Determine whether the DN values ​​of the three pixels at the same position of the cube are monotonically increasing; Step 3-2-2: If it is not monotonically increasing, it is judged as an abnormal pixel and the position of the point is marked; Step 3-2-3: If it is monotonically increasing, solve the slope of the three pixels at the same position according to the relationship y=ax+b, and average the slopes of all N monotonically increasing points, and then calculate the root mean square error σ. If the slope of a pixel at a certain position is and the slope mean If the difference is greater than 3σ, it is defined as an abnormal pixel point; the method for calculating the root mean square error σ refers to the following formula: 。

Images