Noise suppression method and system for black light camera

Abstract

The invention relates to the technical field of image noisy point removal, and discloses a noisy point suppression method and system for a black light camera, and the method comprises the steps: constructing a response rate factor matrix and an intercept factor matrix, carrying out the collection of a working scene image based on the black light camera, obtaining a working scene image, carrying out the correction of the working scene image, and obtaining a noise point suppression result. Performing stripe noise suppression operation on the corrected working scene image to obtain a stripe noise removed image; performing connected domain analysis operation on the binarized image to obtain an independent connected domain set; performing central point diffusion operation on the independent connected domain set to obtain an expanded independent connected domain set; and performing noisy point distinguishing operation on the extended independent connected domain set to obtain a denoised working scene image, and completing noisy point suppression of the black light camera based on the denoised working scene image. According to the invention, stripe noise can be effectively suppressed, defective pixel noise points can be accurately eliminated, and effective targets can be accurately distinguished.

Description

Technical Field

[0001] This invention relates to the field of image noise removal technology, and in particular to a noise suppression method and system for black light cameras. Background Technology

[0002] Blacklight cameras are surveillance cameras capable of producing clear images in extremely low-light environments. Noise suppression refers to the process of eliminating or reducing non-target information interference in an image caused by the physical characteristics of the imaging device, environmental factors, etc., through image processing algorithms and techniques.

[0003] When black light cameras image in low-light environments, due to the influence of detector manufacturing processes and readout circuit structures, two typical types of noise commonly exist in the images: one is stripe non-uniformity noise caused by differences in column amplifier parameters, which manifests as bright and dark stripes in the vertical direction; the other is defective pixel noise generated by defects in the infrared focal plane array, which manifests as isolated bright spots or clusters of tiny bright spots. In existing noise suppression techniques, two-point correction algorithms have fixed parameters and cannot eliminate residual non-uniformity between columns; traditional Top-hat algorithms easily leave bright stripes; and single-point noise denoising algorithms cannot identify noise clusters and are prone to misidentifying targets. Therefore, how to effectively suppress stripe noise, accurately remove defective pixel noise, and accurately distinguish valid targets is an urgent technical problem to be solved. Summary of the Invention

[0004] This invention provides a noise suppression method and a computer-readable storage medium for black light cameras. Its main purpose is to effectively suppress stripe noise, accurately remove defective pixel noise, and accurately distinguish valid targets.

[0005] To achieve the above objectives, the present invention provides a noise suppression method for a black light camera, comprising: The black light camera and blackbody radiation source were identified. The operating temperature range of the camera was determined based on the black light camera. The response factor matrix and intercept factor matrix were constructed based on the operating temperature range of the camera and the blackbody radiation source. Work scene images are acquired using a black light camera, and the work scene images consist of multiple work scene image pixels. Image correction is performed on the work scene image using the response rate factor matrix and the intercept factor matrix to obtain the corrected work scene image; A stripe noise suppression operation is performed on the correction scene image to obtain an image with stripe noise removed; Set a grayscale threshold, and perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image; Perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform a center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. Noise discrimination is performed on the extended independent connected domain set to obtain a denoised working scene image, and noise suppression of the black light camera is completed based on the denoised working scene image.

[0006] Optionally, the construction of the response factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source includes: Extract the upper and lower operating temperature limits from the camera's operating temperature range; Image acquisition operations are performed on the blackbody radiation source based on the upper limit and lower limit of the operating temperature to obtain a calibration image set. The calibration image set includes: low temperature working scene images and high temperature working scene images, wherein the low temperature working scene images include low temperature image pixels. Perform the following operations on each calibration image in the calibration image set: Based on the calibration image, obtain a set of pixel values ​​for multiple columns of images, calculate the average pixel value of each column of image pixel values ​​in the set of multiple columns of image pixel values, and obtain a set of average pixel values, wherein the average pixel value corresponds one-to-one with each column of image pixel values; The pixel average value groups are aggregated to obtain a pixel average value set, which includes: low temperature pixel average value group and high temperature pixel average value group; The correction coefficients are calculated based on the set of pixel average values ​​to obtain the set of response rate factors and the set of intercept factors. Construct response rate factor matrix and intercept factor matrix based on response rate factor set and intercept factor set.

[0007] Optionally, the step of calculating correction coefficients based on a set of pixel average values ​​to obtain a set of response rate factors and a set of intercept factors includes: Low-temperature image pixels are extracted sequentially from low-temperature working scene images, and the low-temperature image pixel values ​​and positions are determined based on the extracted low-temperature image pixels. Based on the pixel position of the low-temperature image, the high-temperature image pixel value, the target low-temperature pixel average value, and the target high-temperature pixel average value are determined from the high-temperature working scene image and the pixel average value set, respectively. The response rate factor is calculated based on the pixel values ​​of the low-temperature image, the pixel values ​​of the high-temperature image, the average value of the target low-temperature pixels, and the average value of the target high-temperature pixels. The intercept factor is calculated based on the response rate factor, the average value of the target high-temperature pixels, and the pixel value of the high-temperature image. The response rate factor and intercept factor are summarized separately to obtain the response rate factor set and intercept factor set.

[0008] Optionally, the formula for calculating the response rate factor is as follows: ; in, Represents the response rate factor. This represents the average value of the target high-temperature pixels. This represents the average value of the target low-temperature pixels. Represents the pixel value of a high-temperature image. This represents the pixel value of a low-temperature image.

[0009] Optionally, the step of performing image correction on the work scene image using the response rate factor matrix and the intercept factor matrix to obtain a corrected work scene image includes: The working scene image pixels are extracted sequentially from the working scene image, the working scene image pixel values ​​are determined based on the working scene image pixels, and the correction response rate factor and calibration intercept factor are determined from the response rate factor matrix and intercept factor matrix respectively based on the working scene image pixel values. The corrected image pixel value is calculated based on the image pixel value of the working scene, the correction response rate factor, and the calibration intercept factor. The calculation formula for the corrected image pixel value is as follows: ; in, This indicates the correction of image pixel values. This represents the corrected response rate factor. Represents the pixel values ​​of the work scene image. Indicates the calibration intercept factor; The pixel values ​​of the corrected images are summarized to obtain the set of corrected image pixel values. Based on the set of corrected image pixel values, the corrected working scene image is identified.

[0010] Optionally, the step of performing stripe noise suppression on the corrected scene image to obtain a stripe noise-removed image includes: The measurement target is identified from the calibration work scene image, the target size is obtained based on the measurement target, and the column structure vector is determined based on the target size; The pixel values ​​of the correction image are extracted sequentially from the correction work scene image. The extracted pixel values ​​of the correction image are used as the pixel values ​​of the image to be eroded. The center element of the vector is identified from the column structure vector. Align the vector center element with the pixel value of the image to be eroded to obtain the column structure covered pixel value set. Based on the column structure vector and the column structure covered pixel value set, the number of column structure elements and the number of covered pixel values ​​are determined. If the number of column structure elements is equal to the number of covered pixel values, then the minimum value extraction operation is performed on the set of covered pixel values ​​of the column structure to obtain the minimum value of the covered pixels. The minimum value of the covered pixels is then used to replace the pixel values ​​of the image to be eroded to obtain the updated image pixel values. If the number of column structure elements is not equal to the number of covered pixel values, then the set of pixel positions to be filled is determined based on the set of covered pixel values ​​of the column structure. The set of pixel positions to be filled includes one or more pixel positions to be filled. The pixel value filling operation is performed on the set of pixel positions to be filled using the preset filling pixel values ​​to obtain the set of filled pixel positions. The updated overlay pixel value set is determined based on the set of filled pixel positions and the column structure overlay pixel value set. The updated image pixel values ​​are obtained based on the updated overlay pixel value set. Summarize the updated image pixel values ​​to obtain the updated image pixel value set, and identify the eroded working scene image based on the updated image pixel value set; A dilation operation is performed on the eroded work scene image to obtain a dilated work scene image. Based on the corrected work scene image and the dilated work scene image, an image with stripe noise removed is identified.

[0011] Optionally, the step of performing a dilation operation on the eroded work scene image to obtain a dilated work scene image includes: The updated image pixel values ​​are extracted sequentially from the eroded working scene image, and the extracted updated image pixel values ​​are used as the pixel values ​​of the image to be expanded; Align the vector center element with the pixel value of the image to be dilated to obtain the column structure neighborhood pixel value set, and determine the number of column neighborhood pixel values ​​based on the column structure neighborhood pixel value set. If the number of column structure elements is equal to the number of column neighbor pixel values, then the maximum value extraction operation is performed on the pixel value set covered by the column structure to obtain the maximum value of the neighbor pixels. The maximum value of the neighbor pixels is then used to replace the pixel values ​​of the image to be dilated to obtain the pixel values ​​of the dilated image. If the number of column structure elements is not equal to the number of pixel values ​​in the column neighborhood, then the pixel values ​​of the expanded image are obtained according to the pixel value set covered by the column structure. Summarize the pixel values ​​of the inflated image to obtain the set of inflated image pixel values, and identify the inflated working scene image based on the set of inflated image pixel values.

[0012] Optionally, the step of performing a centroid diffusion operation on the set of independent connected components to obtain an expanded set of independent connected components includes: Extract independent connected components sequentially from the set of independent connected components; The set of pixel gray values ​​is identified based on the extracted independent connected components. The maximum pixel gray value is extracted from the set of pixel gray values. The position of the maximum pixel is identified based on the maximum pixel gray value. The position of the maximum pixel is used as the candidate center point. The center point diffusion threshold is calculated based on the maximum pixel gray value, and the neighborhood pixel value set is obtained based on the preset neighborhood interval and candidate center points. Extract neighboring pixel values ​​sequentially from the neighboring pixel value set, and compare the extracted neighboring pixel values ​​with the center point diffusion threshold; If the neighboring pixel value is greater than the center point diffusion threshold, the extracted neighboring pixel value is used as the new neighboring pixel value to be assigned, and the new neighboring pixel value to be assigned is used as the candidate center point. The process of obtaining the neighboring pixel value set based on the preset neighboring interval and candidate center point is repeated until the neighboring pixel value is not greater than the center point diffusion threshold. If the neighboring pixel value is not greater than the center point diffusion threshold, then return to the step of sequentially extracting neighboring pixel values ​​from the neighboring pixel value set until all neighboring pixel values ​​in the neighboring pixel value set have been extracted. Summarize the pixel values ​​of the new neighboring regions to obtain multiple pixel values ​​of the new neighboring regions. Add the multiple pixel values ​​of the new neighboring regions to the extracted independent connected regions to obtain the extended independent connected regions. Summarize the extended independent connected components to obtain the extended independent connected component set corresponding to the independent connected component set.

[0013] Optionally, the step of performing noise discrimination on the extended set of independent connected components to obtain the denoised working scene image includes: For each extended independent connected component in the extended independent connected component set, perform the following operation: The connected component pixel set is obtained by expanding the independent connected components, and the number of connected component pixels in the connected component pixel set is calculated. If the number of connected pixels is less than the preset number of standard pixels, the set of connected pixels is taken as the set of noise pixels, and the background value replacement operation is performed on each noise pixel in the noise pixel set to obtain the background pixel set. If the number of pixels in the connected component is not less than the number of standard pixels, then the set of pixels in the connected component is taken as the effective target pixel set. By summing the background pixel set and the effective target pixel set, the denoising working scene image is obtained.

[0014] To achieve the above objectives, the present invention also provides a noise suppression system for a black light camera, comprising: The calibration matrix construction module is used to identify the black light camera and the blackbody radiation source, identify the camera's operating temperature range based on the black light camera, and construct the response factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source. The image stripe noise suppression module is used to acquire work scene images based on a black light camera to obtain work scene images. The work scene images include multiple work scene image pixels. The work scene images are corrected using a response rate factor matrix and an intercept factor matrix to obtain a corrected work scene image. Stripe noise suppression is performed on the corrected work scene image to obtain an image with stripe noise removed. The image connected component analysis module is used to set a grayscale threshold, perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image, perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. The image denoising module is used to perform noise discrimination operations on the extended independent connected domain set to obtain a denoised working scene image, and to complete noise suppression of the black light camera based on the denoised working scene image.

[0015] To address the above problems, the present invention also provides an electronic device, the electronic device comprising: Memory, storing at least one instruction; The processor executes the instructions stored in the memory to implement the noise suppression method for a black light camera described above.

[0016] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the noise suppression method for a black light camera described above.

[0017] To address the problems described in the background art, this invention identifies the blacklight camera and blackbody radiation source, determines the camera's operating temperature range based on the blacklight camera, and constructs a responsivity factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source. This invention uses a two-point linear correction model, establishing a complete linear response range using the upper and lower extreme temperature points of the operating temperature, effectively compensating for fixed-mode noise caused by differences in detector manufacturing processes. The invention acquires working scene images based on the blacklight camera, obtaining working scene images comprising multiple working scene image pixels, and utilizes the responsivity factor... This invention uses a response rate factor matrix and an intercept factor matrix to perform image correction on the working scene image, resulting in a corrected working scene image. This effectively eliminates fixed-pattern noise caused by inconsistent detector pixel responses in black light cameras. The corrected working scene image has a more uniform background grayscale distribution, providing a higher-quality input image for subsequent stripe noise suppression operations and avoiding misjudgments and misprocessing caused by pixel response differences. Stripe noise suppression is then performed on the corrected working scene image to obtain a stripe noise-removed image. This invention, through the correction of the working scene image... For example, stripe noise suppression effectively eliminates column stripe noise caused by differences in the column amplifier parameters of the readout circuit in black light cameras. A grayscale threshold is set, and foreground and background differentiation is performed on the stripe noise-removed image based on this threshold, resulting in a binarized image. This invention simplifies complex grayscale images into binary images, reducing data volume and facilitating subsequent connected component analysis. Connected component analysis is performed on the binarized image to obtain a set of independent connected components. A center-point diffusion operation is then performed on this set of independent connected components to obtain an expanded set of independent connected components. This invention, through connected component analysis, can accurately identify all interconnected bright regions, separating discrete foreground pixels... The organization is defined as a meaningful region object. Simultaneously, the center point diffusion operation, centered on the gray-scale maximum point of each connected component, expands outward using a region growing algorithm, enabling the complete extraction of the entire pixel range of the target or noise. This solves the problem of missing target edges or incomplete noise extraction caused by improper threshold selection in traditional methods. Noise discrimination is performed on the expanded independent connected component set to obtain a denoised working scene image. Based on this denoised working scene image, noise suppression for the black light camera is achieved. This invention achieves precise separation of noise and valid targets by performing noise discrimination on the expanded independent connected component set and judging based on the number of pixels in the connected components. Therefore, this invention can effectively suppress stripe noise, accurately remove defective pixel noise, and accurately distinguish valid targets. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a noise suppression method for a black light camera according to an embodiment of the present invention. Figure 2This is a functional block diagram of a noise suppression system for a black light camera provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device that implements the noise suppression method for a black light camera, according to an embodiment of the present invention.

[0019] Explanation of reference numerals in the attached figures: 1. Electronic device; 10. Processor; 11. Memory; 12. Bus; 100. Noise suppression system for black light cameras; 101. Correction matrix construction module; 102. Image stripe noise suppression module; 103. Image connected component analysis module; 104. Image denoising completion module.

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides a noise suppression method for black light cameras. The execution entity of the noise suppression method for black light cameras includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the noise suppression method for black light cameras can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0023] Reference Figure 1 The diagram shown is a flowchart illustrating a noise suppression method for a black light camera according to an embodiment of the present invention. In this embodiment, the noise suppression method for a black light camera includes: S1. Identify the black light camera and blackbody radiation source. Based on the black light camera, identify the camera's operating temperature range. Based on the camera's operating temperature range and the blackbody radiation source, construct the response rate factor matrix and intercept factor matrix.

[0024] It should be explained that a blacklight camera is a camera capable of achieving clear full-color imaging in extremely low-light environments. A blackbody radiation source is an artificially created device with characteristics close to an ideal blackbody, capable of generating standard infrared radiation signals at a fixed temperature. An ideal blackbody is an idealized model defined in physics, with an absorptivity and emissivity of 1, and its radiated energy depends only on wavelength and temperature, following Planck's law; it does not exist naturally. The camera's operating temperature range is the range of ambient temperatures within which a blacklight camera can stably and reliably achieve its preset imaging functions (including low-light full-color imaging, radiation signal detection, etc.) without performance degradation or damage to its internal core components (such as image sensors, FPGAs, memory, etc.).

[0025] In detail, the construction of the responsivity factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source includes: Extract the upper and lower operating temperature limits from the camera's operating temperature range; Image acquisition operations are performed on the blackbody radiation source based on the upper limit and lower limit of the operating temperature to obtain a calibration image set. The calibration image set includes: low temperature working scene images and high temperature working scene images, wherein the low temperature working scene images include low temperature image pixels. Perform the following operations on each calibration image in the calibration image set: Based on the calibration image, obtain a set of pixel values ​​for multiple columns of images, calculate the average pixel value of each column of image pixel values ​​in the set of multiple columns of image pixel values, and obtain a set of average pixel values, wherein the average pixel value corresponds one-to-one with each column of image pixel values; The pixel average value groups are aggregated to obtain a pixel average value set, which includes: low temperature pixel average value group and high temperature pixel average value group; The correction coefficients are calculated based on the set of pixel average values ​​to obtain the set of response rate factors and the set of intercept factors. Construct response rate factor matrix and intercept factor matrix based on response rate factor set and intercept factor set.

[0026] It should be explained that the upper and lower operating temperature limits are the maximum and minimum temperatures within the operating temperature range of the black light camera, respectively. Low-temperature working scene images are obtained by adjusting the temperature of the blackbody radiation source to match the lower operating temperature limit of the camera, and then capturing images of the blackbody radiation source in this state using the black light camera. High-temperature working scene images are obtained by adjusting the temperature of the blackbody radiation source to match the upper operating temperature limit of the camera, and then capturing images of the blackbody radiation source in this state using the black light camera. A column image pixel value set is the collection of all pixel values ​​corresponding to a specific column of pixels extracted from the calibration image. The pixel average value is the value obtained by performing an arithmetic average of all pixel values ​​in a single column image pixel value set. The pixel average value group is the collection of pixel average values ​​for all columns in the calibration image. The low-temperature pixel average value group is the collection of pixel average values ​​for all columns in the low-temperature working scene image. The high-temperature pixel average value group is the collection of pixel average values ​​for all columns in the high-temperature working scene image. The detailed steps for calculating the correction coefficients based on the pixel average value group set to obtain the response factor set and intercept factor set will be given later and will not be repeated here. The steps for constructing the response rate factor matrix and intercept factor matrix based on the response rate factor set and intercept factor set are as follows: First, take the same spatial pixel position where the low-temperature working scene image and the high-temperature working scene image are completely aligned and the pixel coordinates correspond one-to-one as the reference, and calculate and generate the unique response rate factor and intercept factor under the fixed coordinates pixel by pixel. Then, all the response rate factors obtained by dual-temperature calibration at the same position (i.e., the average value of the target low-temperature pixels and the average value of the target high-temperature pixels) are strictly mapped and filled to the corresponding row and column points of the matrix according to the uniform and fixed pixel row and column coordinates of the response rate factors in the calibration image, and the matrix arrangement is completed to generate a response rate factor matrix that matches the size of the low-temperature working scene image or the high-temperature working scene image. Similarly, the method of constructing the intercept factor matrix is ​​the same as the method of constructing the response rate factor matrix, and will not be repeated here.

[0027] It should be noted that, since the upper and lower operating temperature limits can cover all actual working conditions of the black light camera, a complete and comprehensive temperature benchmark can be provided for the subsequent calibration process of this invention. This avoids the matrix calibration accuracy deviation problem caused by insufficient temperature coverage, ensuring that the finally constructed factor matrix (i.e., the response factor matrix and intercept factor matrix) can adapt to all working states of the camera. At the same time, the blackbody radiation source can output a stable and controllable standard radiation signal, and the radiation intensity corresponding to different set temperatures has fixed traceability characteristics. The acquired images can truly and accurately reflect the radiation response characteristics of the camera under the corresponding temperature conditions. Therefore, this invention is based on the above-mentioned working conditions. Image acquisition operations are performed on the blackbody radiation source at the upper temperature limit and the lower temperature limit of the working temperature to obtain a calibration image set including images of low temperature working scenarios and high temperature working scenarios. Since the original values ​​of single-column pixels are easily affected by imaging noise, which can cause data deviation, the average value of single-column pixels can effectively suppress random noise interference and accurately characterize the average radiation response intensity of the corresponding column of pixels, providing stable and reliable basic data for the calculation of correction coefficients. Therefore, this invention extracts multiple columns of image pixel value sets for each calibration image in the calibration image set, and calculates the pixel average value corresponding to each column of image pixel value set one by one, finally obtaining the pixel average value group matching the single calibration image.

[0028] In detail, the operation of calculating correction coefficients based on the set of pixel average values ​​to obtain a set of response rate factors and a set of intercept factors includes: Low-temperature image pixels are extracted sequentially from low-temperature working scene images, and the low-temperature image pixel values ​​and positions are determined based on the extracted low-temperature image pixels. Based on the pixel position of the low-temperature image, the high-temperature image pixel value, the target low-temperature pixel average value, and the target high-temperature pixel average value are determined from the high-temperature working scene image and the pixel average value set, respectively. The response rate factor is calculated based on the pixel values ​​of the low-temperature image, the pixel values ​​of the high-temperature image, the average value of the target low-temperature pixels, and the average value of the target high-temperature pixels. The intercept factor is calculated based on the response rate factor, the average value of the target high-temperature pixels, and the pixel value of the high-temperature image. The response rate factor and intercept factor are summarized separately to obtain the response rate factor set and intercept factor set.

[0029] It should be explained that low-temperature image pixels are the pixels that constitute the image within the low-temperature working scene image. Low-temperature image pixel values ​​are the pixel values ​​corresponding to the low-temperature image pixels. Low-temperature image pixel positions are the pixel coordinates calibrated within the low-temperature working scene image. High-temperature image pixel values ​​are the pixel values ​​extracted from the low-temperature image pixel positions at the same pixel coordinates in the high-temperature working scene image. The target low-temperature pixel average and target high-temperature pixel average are the low-temperature pixel average and high-temperature pixel average extracted from the low-temperature pixel average group and high-temperature pixel average group, respectively, based on the low-temperature image pixel positions. The detailed steps for calculating the response rate factor based on the low-temperature image pixel values, high-temperature image pixel values, target low-temperature pixel average, and target high-temperature pixel average will be given later and will not be repeated here. The formula for calculating the intercept factor in the steps for calculating the intercept factor based on the response rate factor, target high-temperature pixel average, and high-temperature image pixel values ​​is as follows: ; in, The intercept factor represents the response factor. The response factor characterizes the sensitivity of the camera's image pixel value at the same pixel location to changes in the magnitude and temperature of blackbody radiation energy. It is used to accurately calibrate the fixed proportional relationship between the actual blackbody radiation input and the camera's image pixel output value. The intercept factor is used to compensate for inherent camera hardware bias, optical path noise floor, and fixed system errors, eliminating imaging reference deviations caused by non-radiative factors, and works in conjunction with the response factor to complete the calibration operation. The response factor set is a collection of response factor values. The intercept factor set is a collection of intercept factors.

[0030] In detail, the formula for calculating the response rate factor is as follows: ; in, Represents the response rate factor. This represents the average value of the target high-temperature pixels. This represents the average value of the target low-temperature pixels. Represents the pixel value of a high-temperature image. This represents the pixel value of a low-temperature image.

[0031] It should be explained that the formula for calculating the response rate factor in this invention uses the difference between the average value of the target high-temperature pixels and the average value of the target low-temperature pixels corresponding to the same pixel position as the change in imaging response, and the difference between the pixel values ​​of the high-temperature image and the low-temperature image as the change in actual radiation input. The linear proportionality coefficient is obtained by dividing the two, which characterizes the inherent sensitivity of the camera's imaging output to the change in blackbody radiation temperature and radiation energy under a single pixel. This can eliminate random noise interference in imaging and accurately establish a linear correspondence between the standard blackbody radiation input and the camera's pixel imaging output, thereby obtaining a quantifiable response rate factor, which provides a benchmark coefficient for subsequent calculation of the intercept factor for compensation and for constructing the response rate factor matrix and the intercept factor matrix.

[0032] S2. Acquire work scene images based on a black light camera to obtain work scene images, wherein the work scene images include multiple work scene image pixels.

[0033] It should be explained that the aforementioned acquisition of work scene images based on a black light camera refers to the operation of acquiring images of an actual work scene using a black light camera. A work scene image is an image obtained by the black light camera in an actual deployment environment, capturing images of a real monitored object or area. Work scene image pixels are the pixels that constitute the work scene image.

[0034] S3. Use the response rate factor matrix and intercept factor matrix to perform image correction on the work scene image to obtain the corrected work scene image.

[0035] In detail, the step of using the response rate factor matrix and the intercept factor matrix to perform image correction on the work scene image to obtain a corrected work scene image includes: The working scene image pixels are extracted sequentially from the working scene image, the working scene image pixel values ​​are determined based on the working scene image pixels, and the correction response rate factor and calibration intercept factor are determined from the response rate factor matrix and intercept factor matrix respectively based on the working scene image pixel values. The corrected image pixel value is calculated based on the image pixel value of the working scene, the correction response rate factor, and the calibration intercept factor. The calculation formula for the corrected image pixel value is as follows: ; in, This indicates the correction of image pixel values. This represents the corrected response rate factor. Represents the pixel values ​​of the work scene image. Indicates the calibration intercept factor; The pixel values ​​of the corrected images are summarized to obtain the set of corrected image pixel values. Based on the set of corrected image pixel values, the corrected working scene image is identified.

[0036] It should be explained that the pixel value of the work scene image is the grayscale value corresponding to a single pixel extracted from the actual acquired work scene image. It should also be noted that, because this solution focuses on calibration and correction of blackbody infrared radiation intensity, the camera converts the received radiation energy into a single-channel brightness signal, using grayscale values ​​to represent the radiation intensity, adapting to the computational requirements of the linear correction formula, and avoiding the problem that color three-channel data cannot directly participate in radiation response modeling calculations. The correction response rate factor and calibration intercept factor are the response rate factor and intercept factor extracted from the response rate factor matrix and intercept factor matrix, respectively, based on the corresponding spatial location of the work scene image pixels. The corrected image pixel value is the value obtained by performing a linear operation on the original work scene image pixel values ​​acquired on-site, matching the corresponding pixel location with the correction response rate factor and calibration intercept factor. The corrected image pixel value set is a collection of corrected image pixel values. The corrected work scene image is an image obtained by remapping and stitching the corrected image pixel value set according to the spatial location of the work scene image pixels.

[0037] S4. Perform stripe noise suppression operation on the correction scene image to obtain an image with stripe noise removed.

[0038] Specifically, the step of performing stripe noise suppression on the correction scene image to obtain a stripe noise-removed image includes: The measurement target is identified from the calibration work scene image, the target size is obtained based on the measurement target, and the column structure vector is determined based on the target size; The pixel values ​​of the correction image are extracted sequentially from the correction work scene image. The extracted pixel values ​​of the correction image are used as the pixel values ​​of the image to be eroded. The center element of the vector is identified from the column structure vector. Align the vector center element with the pixel value of the image to be eroded to obtain the column structure covered pixel value set. Based on the column structure vector and the column structure covered pixel value set, the number of column structure elements and the number of covered pixel values ​​are determined. If the number of column structure elements is equal to the number of covered pixel values, then the minimum value extraction operation is performed on the set of covered pixel values ​​of the column structure to obtain the minimum value of the covered pixels. The minimum value of the covered pixels is then used to replace the pixel values ​​of the image to be eroded to obtain the updated image pixel values. If the number of column structure elements is not equal to the number of covered pixel values, then the set of pixel positions to be filled is determined based on the set of covered pixel values ​​of the column structure. The set of pixel positions to be filled includes one or more pixel positions to be filled. The pixel value filling operation is performed on the set of pixel positions to be filled using the preset filling pixel values ​​to obtain the set of filled pixel positions. The updated overlay pixel value set is determined based on the set of filled pixel positions and the column structure overlay pixel value set. The updated image pixel values ​​are obtained based on the updated overlay pixel value set. Summarize the updated image pixel values ​​to obtain the updated image pixel value set, and identify the eroded working scene image based on the updated image pixel value set; A dilation operation is performed on the eroded work scene image to obtain a dilated work scene image. Based on the corrected work scene image and the dilated work scene image, an image with stripe noise removed is identified.

[0039] It should be explained that the measurement target is a reference object used to determine the size of structuring elements in the image of the correction scene. For example, in noise suppression, the measurement target is usually a representative foreground object (such as pedestrians, vehicles, etc.) in the image. The target size is the pixel range occupied by the measurement target in the image space. The target size is represented by the pixel length of the measurement target in the vertical direction (row direction). The column structure vector is the vertical structuring element used for morphological operations, and it is an L×1 column vector, where L is the length of the column structure vector. It should be understood that the length of the column structure vector is greater than the target size to ensure that the opening operation can correctly extract the background rather than the target. For example, the column structure vector is... ,in, Represents a column-oriented structure vector. This indicates the transpose symbol. The pixel value of the image to be eroded is the pixel value of the corrected image extracted point by point from the corrected working scene image. The vector center element is the element at the center of the column structure vector. The alignment operation between the vector center element and the pixel value of the image to be eroded is based on the vector center element, vertically aligning and fitting the entire column structure vector to the pixel position of the current pixel value of the image to be eroded, completing the operation of precise matching and positioning of the structure template and the local pixel region of the image. For example, the corrected working scene image is... The column structure vector is Extract from the third row of the first column of the image of the working scene to be corrected As the pixel value of the image to be eroded, the 1 in the third row of the column structure vector is aligned with the pixel value of the image to be eroded, and the set of pixel values ​​covered by the column structure is {120, 118, 115, 112, 110}.

[0040] Importantly, the column structure covered pixel value set is the set of all corrected image pixel values ​​covered by the column structure vector within the vertical range of the image after alignment. The number of column structure elements is the number of elements in the column structure vector. The number of covered pixel values ​​is the number of pixel values ​​contained in the column structure covered pixel value set. If the number of column structure elements equals the number of covered pixel values, it indicates that the current operation position is within the complete region inside the image of the correction scene, without boundary missing pixels or image truncation. The column structure vector can achieve full and complete coverage, and the standard minimum value erosion operation can be directly performed. The minimum covered pixel value is the smallest column structure covered pixel value in the column structure covered pixel value set. The replacement of the image pixel value to be eroded using the minimum covered pixel value is the operation of replacing the image pixel value to be eroded with the minimum covered pixel value. The updated image pixel value is the new pixel value obtained after minimum value replacement or pixel filling compensation. If the number of column structure elements is not equal to the number of covered pixel values, it indicates that the current operation position is at the edge, corner, or boundary of the measurement target. The column structure vector cannot completely cover the surrounding pixels, and some pixels exceed the range of the correction scene image, resulting in sampling loss. Therefore, the conventional minimum erosion operation cannot be performed directly. The set of pixel positions to be filled is the set of blank positions corresponding to the missing pixels within the coverage area of ​​the structure vector. The fill pixel values ​​are pre-set pixel values ​​used to fill the missing pixel gaps at the boundary of the correction scene image, ensuring a uniform sampling scale of the column structure vector. The fill pixel values ​​are set using the boundary pixel copying method. That is, when the column structure vector exceeds the upper boundary of the correction scene image during the erosion operation, the pixel values ​​of the first row corresponding to the column of the correction scene image are copied as fill values ​​to the pixel positions of the excess portion. When it exceeds the lower boundary of the correction scene image, the pixel values ​​of the last row corresponding to the column of the correction scene image are copied as fill values ​​to the pixel positions of the excess portion.

[0041] It is understandable that the filled pixel position set is the set of pixel coordinates formed after all empty spaces in the unfilled pixel position set have been filled with pixel values. The step of determining the updated covered pixel value set based on the filled pixel position set and the column structure covered pixel value set is the set of pixel values ​​obtained by reintegrating the filled pixel position set with the original column structure covered pixel value set. The method for obtaining updated image pixel values ​​based on the updated covered pixel value set is the same as the method for obtaining updated image pixel values ​​based on the column structure vector and the column structure covered pixel value set, and will not be repeated here. The updated image pixel value set is a set composed of updated image pixel values. The eroded working scene image is the image obtained after performing an erosion operation on the corrected working scene image. The detailed steps for performing a dilation operation on the eroded working scene image to obtain the dilated working scene image will be given later and will not be repeated here. The step of identifying the stripe noise-removed image based on the corrected working scene image and the expanded working scene image is as follows: multiply the pixel values ​​corresponding to the same pixel position in the corrected working scene image and the expanded working scene image one by one, and use the multiplied values ​​as the pixel values ​​of the stripe noise-removed image at the corresponding pixel positions. Repeat this process for all pixels to finally generate the stripe noise-removed image.

[0042] It should be noted that, since vertical stripe noise in the corrected work scene image causes abnormally high pixel brightness at the corresponding location, a column-oriented structural vector that conforms to the stripe direction is used to define the vertical pixel range and extract the minimum pixel value. This directly suppresses the abnormally bright pixels caused by bright stripes, thereby weakening the vertical stripe interference. By relying on a unified vertical sampling method to adapt to the noise distribution pattern, after pixel replacement, excess bright stripes can be removed in the erosion stage. Furthermore, by filling in pixel values ​​to complete the incomplete boundary calculation data, the generated eroded work scene image effectively suppresses stripe defects, laying the foundation for subsequent restoration of normal image details through dilation operations and ultimately obtaining a clean, stripe-free image. Compared with the traditional Top-hat algorithm, the stripe noise suppression operation of this invention uses a column-oriented structural vector specifically designed for noise in the column direction, which can avoid the problem of bright stripe residue. Compared with the column histogram offset correction algorithm, this method not only reduces the non-uniformity of the corrected work scene image but also more effectively changes the non-uniformity within each column of pixels.

[0043] Specifically, the process of performing a dilation operation on the eroded work scene image to obtain a dilated work scene image includes: The updated image pixel values ​​are extracted sequentially from the eroded working scene image, and the extracted updated image pixel values ​​are used as the pixel values ​​of the image to be expanded; Align the vector center element with the pixel value of the image to be dilated to obtain the column structure neighborhood pixel value set, and determine the number of column neighborhood pixel values ​​based on the column structure neighborhood pixel value set. If the number of column structure elements is equal to the number of column neighbor pixel values, then the maximum value extraction operation is performed on the pixel value set covered by the column structure to obtain the maximum value of the neighbor pixels. The maximum value of the neighbor pixels is then used to replace the pixel values ​​of the image to be dilated to obtain the pixel values ​​of the dilated image. If the number of column structure elements is not equal to the number of pixel values ​​in the column neighborhood, then the pixel values ​​of the expanded image are obtained according to the pixel value set covered by the column structure. Summarize the pixel values ​​of the inflated image to obtain the set of inflated image pixel values, and identify the inflated working scene image based on the set of inflated image pixel values.

[0044] It should be explained that the pixel values ​​of the image to be expanded are the updated image pixel values ​​extracted from the eroded working scene image. The column-structured neighborhood pixel value set is the set of all updated image pixel values ​​covered within the vertical neighborhood of the eroded working scene image after alignment with the pixel values ​​of the image to be expanded, using the vector center element as a reference. The number of column-structured neighborhood pixel values ​​is the number of column-structured neighborhood pixel values ​​in the column-structured neighborhood pixel value set. The maximum neighborhood pixel value is the largest column-structured coverage pixel value extracted from the corresponding column-structured neighborhood pixel value set when the column-structured vector can completely cover the vertical pixels of the eroded working scene image. The replacement of the pixel values ​​of the image to be expanded using the maximum neighborhood pixel value is the operation of replacing the pixel values ​​of the image to be expanded with the maximum neighborhood pixel value. The expanded image pixel value is the final pixel value obtained after maximum value replacement processing or edge adaptation. The method of obtaining the expanded image pixel value based on the column-structured coverage pixel value set is the same as the method of obtaining the updated image pixel value based on the column-structured vector and the column-structured coverage pixel value set, and will not be repeated here. The expanded work scene image is an image reconstructed from the pixel value set of the expanded image, according to the coordinates of the pixels in the eroded work scene image.

[0045] It should be noted that since the initial erosion operation suppresses the vertical stripe noise of the corrected scene image by extracting local minimum values, this process will simultaneously reduce the brightness of normal pixels, resulting in darkening of effective imaging details and weakening of vertical texture. Therefore, this invention extracts the maximum value of the pixel value set covered by the column structure and uses a vertical structure template (column structure vector) that matches the stripe direction to extract the brightest pixel in the neighborhood (i.e., the maximum value of the neighborhood pixels). This restores the normal image brightness and edge details that were suppressed during the erosion stage, compensates for the pixel loss caused by noise reduction, weakens the influence of residual dark stripes, and allows the dilation operation to both preserve the stripe removal effect of the initial operation and restore the true and clear image grayscale information, so that the final pixel value conforms to the infrared radiation imaging characteristics of the actual scene.

[0046] S5. Set a grayscale threshold, and perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image.

[0047] It should be explained that the grayscale threshold is set using the Otsu's method. This method can be implemented using existing technology and will not be elaborated upon here. The steps for distinguishing the foreground and background of the image after removing stripe noise based on the grayscale threshold are as follows: The grayscale values ​​of all pixels in the image are read sequentially. All pixels with actual grayscale values ​​greater than or equal to the threshold are uniformly designated as foreground pixels of the target object. These pixels can completely retain the true and effective imaging radiation information and target contour features. Simultaneously, all pixels with actual grayscale values ​​less than the threshold are designated as background pixels containing only environmental interference and no effective target information, thus eliminating redundant low-grayscale invalid image content. The binarized image is obtained by representing each pixel in the image with two grayscale values ​​after the foreground and background distinction operation. It should be noted that foreground pixels are assigned a value of 1 (or 255, representing white), and background pixels are assigned a value of 0 (representing black).

[0048] S6. Perform connected component analysis on the binarized image to obtain a set of independent connected components. Perform center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components.

[0049] It should be explained that the connected component analysis operation on the binarized image is performed using a conventional eight-neighbor or four-neighbor connected component labeling algorithm from the field of image morphology. The set of independent connected components is the collection of all unconnected and independent pixel blocks obtained after the binarized image has been split through connected component analysis.

[0050] In detail, the process of performing a center-point diffusion operation on the set of independent connected components to obtain an expanded set of independent connected components includes: Extract independent connected components sequentially from the set of independent connected components; The set of pixel gray values ​​is identified based on the extracted independent connected components. The maximum pixel gray value is extracted from the set of pixel gray values. The position of the maximum pixel is identified based on the maximum pixel gray value. The position of the maximum pixel is used as the candidate center point. The center point diffusion threshold is calculated based on the maximum pixel gray value, and the neighborhood pixel value set is obtained based on the preset neighborhood interval and candidate center points. Extract neighboring pixel values ​​sequentially from the neighboring pixel value set, and compare the extracted neighboring pixel values ​​with the center point diffusion threshold; If the neighboring pixel value is greater than the center point diffusion threshold, the extracted neighboring pixel value is used as the new neighboring pixel value to be assigned, and the new neighboring pixel value to be assigned is used as the candidate center point. The process of obtaining the neighboring pixel value set based on the preset neighboring interval and candidate center point is repeated until the neighboring pixel value is not greater than the center point diffusion threshold. If the neighboring pixel value is not greater than the center point diffusion threshold, then return to the step of sequentially extracting neighboring pixel values ​​from the neighboring pixel value set until all neighboring pixel values ​​in the neighboring pixel value set have been extracted. Summarize the pixel values ​​of the new neighboring regions to obtain multiple pixel values ​​of the new neighboring regions. Add the multiple pixel values ​​of the new neighboring regions to the extracted independent connected regions to obtain the extended independent connected regions. Summarize the extended independent connected components to obtain the extended independent connected component set corresponding to the independent connected component set.

[0051] It needs to be explained that an independent connected component is a region of a single target contour. The set of pixel grayscale values ​​is the collection of all pixel grayscale values ​​contained within the independent connected component. The maximum pixel grayscale value is the largest pixel grayscale value in the set. The maximum pixel position is the position of the pixel corresponding to the maximum pixel grayscale value within the independent connected component. The candidate center point is the position of the maximum pixel. The calculation of the center point diffusion threshold based on the maximum pixel grayscale value uses 0.25 multiplied by the maximum pixel grayscale value as the center point diffusion threshold. The center point diffusion threshold can both retain effective edge pixels close to the core signal and intercept invalid pixels with excessive grayscale deviations. The neighborhood interval is a pre-defined range for searching the pixel periphery, expanding outward from the candidate center point. For example, the neighborhood interval is 3... 3. The neighborhood pixel value set is the collection of grayscale values ​​of all pixels acquired after defining the search range according to the neighborhood interval, with the candidate center point as the center. A neighborhood pixel value is a pixel value within the neighborhood interval. If a neighborhood pixel value is greater than the center point diffusion threshold, it indicates that the grayscale brightness of the neighborhood pixel is very close to that of the candidate center point (target core region), suggesting a strong correlation between the neighborhood pixel and the imaging signal of the target core region. It is not a cluttered background pixel or an interfering noise pixel, but rather an effective edge or extension pixel of the target itself, meeting the conditions for being added to the original independent connected region. The neighborhood pixel value to be newly assigned is a neighborhood pixel value greater than the center point diffusion threshold. If the neighborhood pixel value is not greater than the center point diffusion threshold, it indicates that the grayscale of the neighborhood pixel is too low and the difference from the candidate center point is too large, belonging to an invalid edge or interfering pixel, and not meeting the conditions for being added to the original independent connected region. Expanding the independent connected region involves adding all the neighborhood pixels to be newly assigned to the original single independent connected region and then re-integrating the resulting region. The extended set of independent connected components is a set consisting of extended independent connected components.

[0052] S7. Perform noise discrimination operation on the extended independent connected domain set to obtain the denoised working scene image, and complete the noise suppression of the black light camera based on the denoised working scene image.

[0053] In detail, the noise discrimination operation on the extended independent connected component set to obtain the denoised working scene image includes: For each extended independent connected component in the extended independent connected component set, perform the following operation: The connected component pixel set is obtained by expanding the independent connected components, and the number of connected component pixels in the connected component pixel set is calculated. If the number of connected pixels is less than the preset number of standard pixels, the set of connected pixels is taken as the set of noise pixels, and the background value replacement operation is performed on each noise pixel in the noise pixel set to obtain the background pixel set. If the number of pixels in the connected component is not less than the number of standard pixels, then the set of pixels in the connected component is taken as the effective target pixel set. By summing the background pixel set and the effective target pixel set, the denoising working scene image is obtained.

[0054] It should be explained that the connected component pixel set is the collection of all pixels contained within the extended independent connected components. The number of connected component pixels is the number of connected component pixels in the connected component pixel set. If the number of connected component pixels is less than the preset standard number of pixels, it indicates that the pixel area occupied by the current extended independent connected components is too small, failing to meet the minimum size requirement of the actual target being measured. This is considered as scattered white spots, isolated noise, or residual interference generated during imaging, and is not a valid target detection. The standard number of pixels is a pre-set, fixed threshold number of pixels used as the area judgment criterion to distinguish between the real target area and the small noise area. The standard pixel count is set as follows: First, calibration is performed based on target imaging features. Representative typical targets (such as pedestrians, vehicles, or stars) from actual application scenarios of black light cameras are selected. A large number of actual working scene images containing these typical targets are collected. Target regions are extracted through manual annotation. The pixel area occupied by each typical target in the actual working scene images is counted. The minimum pixel area value of each type of typical target in all actual working scene images is calculated. For example, the minimum area for pedestrians is 15 pixels, the minimum area for vehicles is 25 pixels, and the minimum area for stars is 4 pixels. The minimum value among the minimum areas of each type of target is taken as the reference upper limit of the standard pixel count. Secondly, calibration is performed by combining noise statistical characteristics. The black light camera is pointed at a uniform blackbody radiation source, and no less than 50 frames of uniform background images are collected. The pixel area distribution of defective pixel noise in the uniform background images is statistically analyzed. The occurrence frequency of isolated noise points (area of ​​1 pixel) and small noise clusters (area of ​​2 to 3 pixels) is recorded. The upper limit of the noise area (i.e., 3 pixels) is taken as the lower limit reference of the standard number of pixels. Finally, combining the above two statistical results, between the upper limit of the noise area (3 pixels) and the lower limit of the minimum target area (4 pixels), based on debugging experience and the requirements of noise removal rate and target retention rate in actual application, the standard number of pixels is set to 4 pixels.

[0055] Importantly, the noisy pixel set is a set of connected pixels where the number of connected pixels is less than the standard pixel set. The background value replacement operation for each noisy pixel in the noisy pixel set involves replacing the grayscale value of each pixel in the noisy pixel set with a background value. The background pixel set is the set of pixels after the background value replacement operation is performed on each noisy pixel in the noisy pixel set, replacing the pixel values ​​of the noisy pixels with background values. If the number of connected pixels is not less than the standard pixel set, it means that the pixel area of ​​the current extended independent connected region meets the minimum size requirement, the region size conforms to the imaging characteristics of the actual measured target, it does not belong to scattered interference, and it is a valid imaging region that needs to be retained. The valid target pixel set is a set of connected pixels where the number of connected pixels is not less than the standard pixel set. The denoised scene image is the image obtained by re-integrating and summarizing all background pixels that have undergone background replacement processing and all retained valid target pixels according to the positional relationship of the binarized image.

[0056] To address the problems described in the background art, this invention identifies the blacklight camera and blackbody radiation source, determines the camera's operating temperature range based on the blacklight camera, and constructs a responsivity factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source. This invention uses a two-point linear correction model, establishing a complete linear response range using the upper and lower extreme temperature points of the operating temperature, effectively compensating for fixed-mode noise caused by differences in detector manufacturing processes. The invention acquires working scene images based on the blacklight camera, obtaining working scene images comprising multiple working scene image pixels, and utilizes the responsivity factor... This invention uses a response rate factor matrix and an intercept factor matrix to perform image correction on the working scene image, resulting in a corrected working scene image. This effectively eliminates fixed-pattern noise caused by inconsistent detector pixel responses in black light cameras. The corrected working scene image has a more uniform background grayscale distribution, providing a higher-quality input image for subsequent stripe noise suppression operations and avoiding misjudgments and misprocessing caused by pixel response differences. Stripe noise suppression is then performed on the corrected working scene image to obtain a stripe noise-removed image. This invention, through the correction of the working scene image... For example, stripe noise suppression effectively eliminates column stripe noise caused by differences in the column amplifier parameters of the readout circuit in black light cameras. A grayscale threshold is set, and foreground and background differentiation is performed on the stripe noise-removed image based on this threshold, resulting in a binarized image. This invention simplifies complex grayscale images into binary images, reducing data volume and facilitating subsequent connected component analysis. Connected component analysis is performed on the binarized image to obtain a set of independent connected components. A center-point diffusion operation is then performed on this set of independent connected components to obtain an expanded set of independent connected components. This invention, through connected component analysis, can accurately identify all interconnected bright regions, separating discrete foreground pixels... The organization is defined as a meaningful region object. Simultaneously, the center point diffusion operation, centered on the gray-scale maximum point of each connected component, expands outward using a region growing algorithm, enabling the complete extraction of the entire pixel range of the target or noise. This solves the problem of missing target edges or incomplete noise extraction caused by improper threshold selection in traditional methods. Noise discrimination is performed on the expanded independent connected component set to obtain a denoised working scene image. Based on this denoised working scene image, noise suppression for the black light camera is achieved. This invention achieves precise separation of noise and valid targets by performing noise discrimination on the expanded independent connected component set and judging based on the number of pixels in the connected components. Therefore, this invention can effectively suppress stripe noise, accurately remove defective pixel noise, and accurately distinguish valid targets.

[0057] like Figure 2 The diagram shown is a functional block diagram of a noise suppression system for a black light camera provided in an embodiment of the present invention.

[0058] The noise suppression system 100 for a black light camera described in this invention can be installed in an electronic device. Depending on the functions implemented, the noise suppression system 100 for a black light camera may include a correction matrix construction module 101, an image stripe noise suppression module 102, an image connected component analysis module 103, and an image denoising completion module 104. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device. The correction matrix construction module 101 is used to identify the black light camera and the black body radiation source, identify the camera's operating temperature range based on the black light camera, and construct a response rate factor matrix and an intercept factor matrix based on the camera's operating temperature range and the black body radiation source. The image stripe noise suppression module 102 is used to acquire work scene images based on a black light camera to obtain a work scene image. The work scene image includes multiple work scene image pixels. The work scene image is corrected using a response rate factor matrix and an intercept factor matrix to obtain a corrected work scene image. Stripe noise suppression is performed on the corrected work scene image to obtain an image with stripe noise removed. The image connected component analysis module 103 is used to set a grayscale threshold, perform a foreground-background distinction operation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image, perform a connected component analysis operation on the binarized image to obtain a set of independent connected components, and perform a center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. The image denoising module 104 is used to perform noise discrimination operation on the extended independent connected domain set to obtain a denoised working scene image, and to complete noise suppression of the black light camera based on the denoised working scene image.

[0059] In detail, the modules in the noise suppression system 100 for a black light camera described in this embodiment of the invention employ the same methods as described above during use. Figure 1 The method used is the same as the noise suppression method for black light cameras described in the previous section, and can produce the same technical effect, so it will not be repeated here.

[0060] like Figure 3 The diagram shown is a structural schematic of an electronic device for implementing a noise suppression method for a black light camera, according to an embodiment of the present invention.

[0061] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a noise suppression method program for a black light camera.

[0062] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as code for a noise suppression method program for a black light camera, but also to temporarily store data that has been output or will be output.

[0063] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., noise suppression methods for black light cameras) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0064] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0065] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0066] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power the various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0067] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0068] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0069] The noise suppression method program for the black light camera stored in the memory 11 of the electronic device 1 is a combination of multiple instructions, which, when run in the processor 10, can achieve the following: The black light camera and blackbody radiation source were identified. The operating temperature range of the camera was determined based on the black light camera. The response factor matrix and intercept factor matrix were constructed based on the operating temperature range of the camera and the blackbody radiation source. Work scene images are acquired using a black light camera, and the work scene images consist of multiple work scene image pixels. Image correction is performed on the work scene image using the response rate factor matrix and the intercept factor matrix to obtain the corrected work scene image; A stripe noise suppression operation is performed on the correction scene image to obtain an image with stripe noise removed; Set a grayscale threshold, and perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image; Perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform a center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. Noise discrimination is performed on the extended independent connected domain set to obtain a denoised working scene image, and noise suppression of the black light camera is completed based on the denoised working scene image.

[0070] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0071] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0072] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following: The black light camera and blackbody radiation source were identified. The operating temperature range of the camera was determined based on the black light camera. The response factor matrix and intercept factor matrix were constructed based on the operating temperature range of the camera and the blackbody radiation source. Work scene images are acquired using a black light camera, and the work scene images consist of multiple work scene image pixels. Image correction is performed on the work scene image using the response rate factor matrix and the intercept factor matrix to obtain the corrected work scene image; A stripe noise suppression operation is performed on the correction scene image to obtain an image with stripe noise removed; Set a grayscale threshold, and perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image; Perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform a center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. Noise discrimination is performed on the extended independent connected domain set to obtain a denoised working scene image, and noise suppression of the black light camera is completed based on the denoised working scene image.

[0073] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0074] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0075] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0076] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0077] 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 it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A noise suppression method for a black light camera, characterized in that, The method includes: The black light camera and blackbody radiation source were identified. The operating temperature range of the camera was determined based on the black light camera. The response factor matrix and intercept factor matrix were constructed based on the operating temperature range of the camera and the blackbody radiation source. Work scene images are acquired using a black light camera, and the work scene images consist of multiple work scene image pixels. Image correction is performed on the work scene image using the response rate factor matrix and the intercept factor matrix to obtain the corrected work scene image; A stripe noise suppression operation is performed on the correction scene image to obtain an image with stripe noise removed; Set a grayscale threshold, and perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image; Perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform a center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. Noise discrimination is performed on the extended independent connected domain set to obtain a denoised working scene image, and noise suppression of the black light camera is completed based on the denoised working scene image.

2. The noise suppression method for a black light camera as described in claim 1, characterized in that, The construction of the response factor matrix and intercept factor matrix based on the camera's operating temperature range and blackbody radiation source includes: Extract the upper and lower operating temperature limits from the camera's operating temperature range; Image acquisition operations are performed on the blackbody radiation source based on the upper limit and lower limit of the operating temperature to obtain a calibration image set. The calibration image set includes: low temperature working scene images and high temperature working scene images, wherein the low temperature working scene images include low temperature image pixels. Perform the following operations on each calibration image in the calibration image set: Based on the calibration image, obtain a set of pixel values ​​for multiple columns of images, calculate the average pixel value of each column of image pixel values ​​in the set of multiple columns of image pixel values, and obtain a set of average pixel values, wherein the average pixel value corresponds one-to-one with each column of image pixel values; The pixel average value groups are aggregated to obtain a pixel average value set, which includes: low temperature pixel average value group and high temperature pixel average value group; The correction coefficients are calculated based on the set of pixel average values ​​to obtain the set of response rate factors and the set of intercept factors. Construct response rate factor matrix and intercept factor matrix based on response rate factor set and intercept factor set.

3. The noise suppression method for a black light camera as described in claim 2, characterized in that, The operation of calculating correction coefficients based on the pixel average value set yields a response rate factor set and an intercept factor set, including: Low-temperature image pixels are extracted sequentially from low-temperature working scene images, and the low-temperature image pixel values ​​and positions are determined based on the extracted low-temperature image pixels. Based on the pixel position of the low-temperature image, the high-temperature image pixel value, the target low-temperature pixel average value, and the target high-temperature pixel average value are determined from the high-temperature working scene image and the pixel average value set, respectively. The response rate factor is calculated based on the pixel values ​​of the low-temperature image, the pixel values ​​of the high-temperature image, the average value of the target low-temperature pixels, and the average value of the target high-temperature pixels. The intercept factor is calculated based on the response rate factor, the average value of the target high-temperature pixels, and the pixel value of the high-temperature image. The response rate factor and intercept factor are summarized separately to obtain the response rate factor set and intercept factor set.

4. The noise suppression method for a black light camera as described in claim 3, characterized in that, The formula for calculating the response rate factor is as follows: ； in, Represents the response rate factor. This represents the average value of the target high-temperature pixels. This represents the average value of the target low-temperature pixels. Represents the pixel value of a high-temperature image. This represents the pixel value of a low-temperature image.

5. The noise suppression method for a black light camera as described in claim 4, characterized in that, The process of using the response rate factor matrix and intercept factor matrix to perform image correction on the work scene image to obtain a corrected work scene image includes: The working scene image pixels are extracted sequentially from the working scene image, the working scene image pixel values ​​are determined based on the working scene image pixels, and the corrected response rate factor and the calibration intercept factor are determined from the response rate factor matrix and the intercept factor matrix respectively based on the working scene image pixel values. The corrected image pixel value is calculated based on the image pixel value of the working scene, the correction response rate factor, and the calibration intercept factor. The calculation formula for the corrected image pixel value is as follows: ； in, This indicates the correction of image pixel values. This represents the corrected response rate factor. Represents the pixel values ​​of the work scene image. Indicates the calibration intercept factor; The pixel values ​​of the corrected images are summarized to obtain the set of corrected image pixel values. Based on the set of corrected image pixel values, the corrected working scene image is identified.

6. The noise suppression method for a black light camera as described in claim 5, characterized in that, The step of performing stripe noise suppression on the correction scene image to obtain a stripe noise-removed image includes: The measurement target is identified from the calibration work scene image, the target size is obtained based on the measurement target, and the column structure vector is determined based on the target size; The pixel values ​​of the correction image are extracted sequentially from the correction work scene image. The extracted pixel values ​​of the correction image are used as the pixel values ​​of the image to be eroded. The center element of the vector is identified from the column structure vector. Align the vector center element with the pixel value of the image to be eroded to obtain the column structure covered pixel value set. Based on the column structure vector and the column structure covered pixel value set, the number of column structure elements and the number of covered pixel values ​​are determined. If the number of column structure elements is equal to the number of covered pixel values, then the minimum value extraction operation is performed on the set of covered pixel values ​​of the column structure to obtain the minimum value of the covered pixels. The minimum value of the covered pixels is then used to replace the pixel values ​​of the image to be eroded to obtain the updated image pixel values. If the number of column structure elements is not equal to the number of covered pixel values, then the set of pixel positions to be filled is determined based on the set of covered pixel values ​​of the column structure, wherein the set of pixel positions to be filled includes one or more pixel positions to be filled. The pixel value filling operation is performed on the set of pixel positions to be filled using the preset filling pixel values ​​to obtain the set of filled pixel positions. The updated overlay pixel value set is determined based on the set of filled pixel positions and the column structure overlay pixel value set. The updated image pixel values ​​are obtained based on the updated overlay pixel value set. Summarize the updated image pixel values ​​to obtain the updated image pixel value set, and identify the eroded working scene image based on the updated image pixel value set; A dilation operation is performed on the eroded work scene image to obtain a dilated work scene image. Based on the corrected work scene image and the dilated work scene image, an image with stripe noise removed is identified.

7. The noise suppression method for a black light camera as described in claim 6, characterized in that, The process of performing a dilation operation on the eroded work scene image to obtain a dilated work scene image includes: The updated image pixel values ​​are extracted sequentially from the eroded working scene image, and the extracted updated image pixel values ​​are used as the pixel values ​​of the image to be expanded; Align the vector center element with the pixel value of the image to be dilated to obtain the column structure neighborhood pixel value set, and determine the number of column neighborhood pixel values ​​based on the column structure neighborhood pixel value set. If the number of column structure elements is equal to the number of column neighbor pixel values, then the maximum value extraction operation is performed on the pixel value set covered by the column structure to obtain the maximum value of the neighbor pixels. The maximum value of the neighbor pixels is then used to replace the pixel values ​​of the image to be dilated to obtain the pixel values ​​of the dilated image. If the number of column structure elements is not equal to the number of pixel values ​​in the column neighborhood, then the pixel values ​​of the expanded image are obtained according to the pixel value set covered by the column structure. Summarize the pixel values ​​of the inflated image to obtain the set of inflated image pixel values, and identify the inflated working scene image based on the set of inflated image pixel values.

8. The noise suppression method for a black light camera as described in claim 7, characterized in that, The process of performing a centroid diffusion operation on a set of independent connected components to obtain an expanded set of independent connected components includes: Extract independent connected components sequentially from the set of independent connected components; The set of pixel gray values ​​is identified based on the extracted independent connected components. The maximum pixel gray value is extracted from the set of pixel gray values. The position of the maximum pixel is identified based on the maximum pixel gray value. The position of the maximum pixel is used as the candidate center point. The center point diffusion threshold is calculated based on the maximum pixel gray value, and the neighborhood pixel value set is obtained based on the preset neighborhood interval and candidate center points. Extract neighboring pixel values ​​sequentially from the neighboring pixel value set, and compare the extracted neighboring pixel values ​​with the center point diffusion threshold; If the neighboring pixel value is greater than the center point diffusion threshold, the extracted neighboring pixel value is used as the new neighboring pixel value to be assigned, and the new neighboring pixel value to be assigned is used as the candidate center point. The process of obtaining the neighboring pixel value set based on the preset neighboring interval and candidate center point is repeated until the neighboring pixel value is not greater than the center point diffusion threshold. If the neighboring pixel value is not greater than the center point diffusion threshold, then return to the step of sequentially extracting neighboring pixel values ​​from the neighboring pixel value set until all neighboring pixel values ​​in the neighboring pixel value set have been extracted. Summarize the pixel values ​​of the new neighboring regions to obtain multiple pixel values ​​of the new neighboring regions. Add the multiple pixel values ​​of the new neighboring regions to the extracted independent connected regions to obtain the extended independent connected regions. Summarize the extended independent connected components to obtain the extended independent connected component set corresponding to the independent connected component set.

9. The noise suppression method for a black light camera as described in claim 8, characterized in that, The step of performing noise discrimination on the extended independent connected component set to obtain the denoised scene image includes: For each extended independent connected component in the extended independent connected component set, perform the following operation: Based on the extended independent connected components, obtain the connected component pixel set and calculate the number of connected component pixels in the connected component pixel set; If the number of connected pixels is less than the preset number of standard pixels, the set of connected pixels is taken as the set of noise pixels, and the background value replacement operation is performed on each noise pixel in the noise pixel set to obtain the background pixel set. If the number of pixels in the connected component is not less than the number of standard pixels, then the set of pixels in the connected component is taken as the effective target pixel set. By summing the background pixel set and the effective target pixel set, the denoising working scene image is obtained.

10. A noise suppression system for a black light camera, characterized in that, The system includes: The calibration matrix construction module is used to identify the black light camera and the blackbody radiation source, identify the camera's operating temperature range based on the black light camera, and construct the response factor matrix and intercept factor matrix based on the camera's operating temperature range and the blackbody radiation source. The image stripe noise suppression module is used to acquire work scene images based on a black light camera to obtain work scene images. The work scene images include multiple work scene image pixels. The work scene images are corrected using a response rate factor matrix and an intercept factor matrix to obtain a corrected work scene image. Stripe noise suppression is performed on the corrected work scene image to obtain an image with stripe noise removed. The image connected component analysis module is used to set a grayscale threshold, perform foreground and background differentiation on the stripe noise-removed image based on the grayscale threshold to obtain a binarized image, perform connected component analysis on the binarized image to obtain a set of independent connected components, and perform center point diffusion operation on the set of independent connected components to obtain an extended set of independent connected components. The image denoising module is used to perform noise discrimination operations on the extended independent connected domain set to obtain a denoised working scene image, and to complete noise suppression of the black light camera based on the denoised working scene image.

Images