阅读说明：本技术 一种星敏感器星点像斑检测方法及系统 (Star sensor star point image spot detection method and system ) 是由 杨飞 于 2021-06-16 设计创作，主要内容包括：本发明公开了一种星敏感器星点像斑检测方法及系统,在星敏感器所成图像上进行扫描,首先检测当前扫描像素灰度邻域灰度最大值的像素位置,其次计算当前扫描像素邻域背景的灰度均值和灰度标准差值,进而得到自适应阈值,最后根据自适应阈值分割星点像斑灰度图像,满足一系列检测标准情况后则检测出星点像斑,逐行或者逐列扫描完所有像素位置后,检测出所有的星点像斑。本发明的优点是：实现简单,可以获得精准的局部阈值,避免整帧图像背景灰度分布不均匀造成的检测阈值失效造成的漏检和虚警情况可以准确筛除背景图像中的孤立噪声,并且检测出满足要求的星点像斑,获得更精准的分割阈值,使用的星点像斑检测标准逻辑简单,利于FPGA逻辑实现。(The invention discloses a star point image spot detection method and a star point image spot detection system for a star sensor, wherein scanning is carried out on an image formed by the star sensor, the pixel position of the maximum value of the gray level neighborhood of a currently scanned pixel is detected firstly, the gray level mean value and the gray level standard difference value of the background of the neighborhood of the currently scanned pixel are calculated, so that an adaptive threshold value is obtained, the gray level image of the star point image spot is segmented according to the adaptive threshold value, the star point image spot is detected after a series of detection standard conditions are met, and all the star point image spots are detected after all the pixel positions are scanned line by line or line by line. The invention has the advantages that: the method is simple to realize, can obtain an accurate local threshold, can accurately screen out isolated noise in the background image under the conditions of missing detection and false alarm caused by detection threshold failure due to uneven background gray level distribution of the whole frame image, can detect out the star spot meeting the requirements, obtains a more accurate segmentation threshold, has simple detection standard logic of the used star spot, and is beneficial to the realization of FPGA logic.)

1. A star sensor star point image spot detection method is characterized by comprising the following steps:

acquiring a frame of star spot gray image, wherein the pixel resolution of the star spot gray image is M multiplied by N, the number of lines is M, the number of columns is N, scanning is performed line by line or column by column, the size of a scanning window is set to be qxq, q is an odd number, 0< q is less than or equal to 7, the size of a threshold calculation window is N multiplied by N neighborhood, N is an odd number, 15> N > q, and the scanning range is as follows:

for each pixel position in the scanning range, judging whether the gray value of the current pixel position is the maximum value of the gray values of q multiplied by q neighborhood pixels, if so, turning to the next step, otherwise, continuing to scan the next pixel position in the scanning range;

when the gray value of the current pixel position is determined to be the maximum value of the gray value of the q multiplied by q neighborhood pixels, the pixel position where the current scanning position is the maximum value of the gray value of the q multiplied by q neighborhood pixels is taken as the center to obtain the gray value of the pixel covered by the n multiplied by n window;

calculating the gray level mean value and the gray level standard difference value of all pixels in the nxn window according to the gray level values of the pixels covered by the nxn window, and then obtaining an adaptive threshold value by utilizing the gray level mean value and the gray level standard difference value;

and counting the number of pixels with the pixel gray values larger than the adaptive threshold in the field covered by the q multiplied by q pixels, if the number of the pixels is not smaller than the detection threshold set by a user, considering the scanned position as the star spot position, and if not, considering the scanned position as the star spot position, and finishing the scanning of the current pixel position.

2. The method for detecting the image spot of the star point of the star sensor as claimed in claim 1, wherein the scanning line by line or column by column comprises: and scanning the image from the initial pixel position to the end pixel position line by line or line by line in the scanning range by utilizing the scanning window to obtain the gray value of the current scanning pixel.

3. The method for detecting the star point image spot of the star sensor according to claim 2, wherein the step of judging the gray value of the current pixel position further comprises the steps of: and detecting the pixel position of the maximum gray value of the gray neighborhood of the currently scanned pixel according to the gray value of the currently scanned pixel.

4. The method for detecting the star point image spots of the star sensor according to claim 1, further comprising segmenting the gray scale image of the star point image spots according to the adaptive threshold.

5. The method for detecting the star point image spots of the star sensor as claimed in claim 1, further comprising: and continuing to scan the next pixel position in the scanning range until traversing each pixel position of the star spot gray scale image in the scanning range.

6. A star sensor star point image spot detection system is characterized by comprising:

the analysis unit is used for obtaining a frame of star spot gray image, the pixel resolution of the star spot gray image is M multiplied by N, the number of lines is M, the number of columns is N, the star spot gray image is scanned line by line or column by column, the size of a scanning window is set to be q multiplied by q, q is an odd number, 0< q is less than or equal to 7, the size of a threshold calculation window is N multiplied by N neighborhood, N is an odd number, 15> N > q, and the scanning range is as follows:

for each pixel position in the scanning range, judging whether the gray value of the current pixel position is the maximum value of the gray values of q multiplied by q neighborhood pixels, if so, turning to the next step, otherwise, continuing to scan the next pixel position in the scanning range;

the acquisition unit is used for acquiring the pixel gray value covered by the n multiplied by n window by taking the pixel position of which the current scanning position is the maximum value of the gray value of the q multiplied by q neighborhood pixels as the center when the gray value of the current pixel position is determined to be the maximum value of the gray value of the q multiplied by q neighborhood pixels;

the calculation unit is used for calculating the gray level mean value and the gray level standard difference value of all pixels in the nxn window according to the gray level values of the pixels covered by the nxn window, and then obtaining the self-adaptive threshold value by utilizing the gray level mean value and the gray level standard difference value;

and the determining unit is used for counting the number of pixels with pixel gray values larger than the adaptive threshold in the field covered by the qxq pixels, if the number of the pixels is not smaller than the detection threshold set by the user, the scanned position is considered as the star spot position, otherwise, the scanned position is not considered as the star spot position, and the scanning of the current pixel position is finished.

7. The star sensor star spot image detection system of claim 6, wherein said analysis unit comprises a scanning module for scanning said image from a start pixel position to an end pixel position line by line or column by column within said scanning range by using said scanning window to obtain a gray value of a currently scanned pixel.

8. The star sensor star spot image detection system of claim 7, wherein the analysis unit further comprises a detection module for detecting a pixel position of a maximum gray level in a gray level neighborhood of the currently scanned pixel according to the gray level of the currently scanned pixel.

9. The star sensor star spot detection system of claim 6 wherein said computing unit further comprises a segmentation module for segmenting said star spot gray scale image according to said adaptive threshold.

10. The star sensor star spot detection system of claim 6, further comprising a traversal unit for continuing to scan the next pixel position within the scan range until traversing each pixel position of the star spot gray scale image within the scan range.

Technical Field

The invention belongs to the technical field of astronomical navigation, and particularly relates to a star point image spot detection method and system of a star sensor.

Background

The star sensor is one of astronomical navigation technologies, and the vector direction of an optical axis of the star sensor under a celestial coordinate system is solved by shooting and measuring the star sky of a fixed star and combining the position information of the fixed star in the celestial coordinate system, so that the attitude information of a flying platform where the star sensor is located is obtained. The star sensor firstly detects star point image spots from a shot star image, only allows a small amount of false alarm targets to exist, and avoids the influence of large-size spot targets such as uneven image background gray scale, moon, flying objects and the like.

According to the traditional star spot detection method, a global segmentation threshold is calculated by using a whole frame image, and star spot detection is completed by using connected domain detection. When the global segmentation threshold is used for processing a star map with uneven background gray level, the global segmentation threshold is not the most accurate segmentation threshold for star spot image spots at different positions, and particularly, the threshold cannot be adaptively improved when a large spot target such as a moon or a flying object is detected, so that false fright and missed detection are caused. On the other hand, the traditional connected domain detection algorithm is relatively complex in logic and relatively large in calculation amount, and is not beneficial to FPGA logic realization.

Disclosure of Invention

The invention aims to provide a star point image spot detection method of a star sensor, which solves the problem of star point image spot detection of false startle and missed detection caused by low computation complexity and low signal-to-noise ratio and incapability of improving a threshold value in a self-adaption mode in the prior art.

In view of the above, the present invention provides a method for detecting star point image spots of a star sensor, which is characterized by comprising:

acquiring a frame of star spot gray image, wherein the pixel resolution of the star spot gray image is M multiplied by N, the number of lines is M, the number of columns is N, scanning is performed line by line or column by column, the size of a scanning window is set to be qxq, q is an odd number, 0< q is less than or equal to 7, the size of a threshold calculation window is N multiplied by N neighborhood, N is an odd number, 15> N > q, and the scanning range is as follows:

for each pixel position in the scanning range, judging whether the gray value of the current pixel position is the maximum value of the gray values of q multiplied by q neighborhood pixels, if so, turning to the next step, otherwise, continuing to scan the next pixel position in the scanning range;

when the gray value of the current pixel position is determined to be the maximum value of the gray value of the q multiplied by q neighborhood pixels, the pixel position where the current scanning position is the maximum value of the gray value of the q multiplied by q neighborhood pixels is taken as the center to obtain the gray value of the pixel covered by the n multiplied by n window;

calculating the gray level mean value and the gray level standard difference value of all pixels in the nxn window according to the gray level values of the pixels covered by the nxn window, and then obtaining an adaptive threshold value by utilizing the gray level mean value and the gray level standard difference value;

and counting the number of pixels with the pixel gray values larger than the adaptive threshold in the field covered by the q multiplied by q pixels, if the number of the pixels is not smaller than the detection threshold set by a user, considering the scanned position as the star spot position, and if not, considering the scanned position as the star spot position, and finishing the scanning of the current pixel position.

Further, the row-by-row or column-by-column scanning includes: and scanning the image from the initial pixel position to the end pixel position line by line or line by line in the scanning range by utilizing the scanning window to obtain the gray value of the current scanning pixel.

Further, the determining the gray scale value of the current pixel position further includes: and detecting the pixel position of the maximum gray value of the gray neighborhood of the currently scanned pixel according to the gray value of the currently scanned pixel.

Further, the method further comprises the step of segmenting the star spot gray level image according to the adaptive threshold value.

Further, still include: and continuing to scan the next pixel position in the scanning range until traversing each pixel position of the star spot gray scale image in the scanning range.

Another object of the present invention is to provide a star point image spot detection system of a star sensor, comprising:

the analysis unit is used for obtaining a frame of star spot gray image, the pixel resolution of the star spot gray image is M multiplied by N, the number of lines is M, the number of columns is N, the star spot gray image is scanned line by line or column by column, the size of a scanning window is set to be q multiplied by q, q is an odd number, 0< q is less than or equal to 7, the size of a threshold calculation window is N multiplied by N neighborhood, N is an odd number, 15> N > q, and the scanning range is as follows:

for each pixel position in the scanning range, judging whether the gray value of the current pixel position is the maximum value of the gray values of q multiplied by q neighborhood pixels, if so, turning to the next step, otherwise, continuing to scan the next pixel position in the scanning range;

the acquisition unit is used for acquiring the pixel gray value covered by the n multiplied by n window by taking the pixel position of which the current scanning position is the maximum value of the gray value of the q multiplied by q neighborhood pixels as the center when the gray value of the current pixel position is determined to be the maximum value of the gray value of the q multiplied by q neighborhood pixels;

the calculation unit is used for calculating the gray level mean value and the gray level standard difference value of all pixels in the nxn window according to the gray level values of the pixels covered by the nxn window, and then obtaining the self-adaptive threshold value by utilizing the gray level mean value and the gray level standard difference value;

and the determining unit is used for counting the number of pixels with pixel gray values larger than the adaptive threshold in the field covered by the qxq pixels, if the number of the pixels is not smaller than the detection threshold set by the user, the scanned position is considered as the star spot position, otherwise, the scanned position is not considered as the star spot position, and the scanning of the current pixel position is finished.

Further, the analysis unit includes a scanning module, and scans the image from the start pixel position to the end pixel position line by line or column by column within the scanning range by using the scanning window, so as to obtain the gray value of the current scanning pixel.

Further, the analysis unit further includes a detection module, configured to detect a pixel position of a maximum value of a gray level in a gray level neighborhood of the currently scanned pixel according to the gray level of the currently scanned pixel.

Further, the computing unit further comprises a segmentation module for segmenting the star spot gray scale image according to the adaptive threshold.

And further, the traversing unit is used for continuously scanning the next pixel position in the scanning range until each pixel position of the star spot gray scale image in the scanning range is traversed.

The invention achieves the following significant beneficial effects:

the realization is simple, include: firstly, detecting the pixel position of the maximum value of the gray level neighborhood of the current scanning pixel, secondly, calculating the gray level mean value and the gray level standard difference value of the background of the neighborhood of the current scanning pixel so as to obtain an adaptive threshold, and finally, segmenting the gray level image of the star spot according to the adaptive threshold, and considering the star spot image as the detected star spot after meeting a series of detection standards. After all pixel positions are scanned line by line or column by column, all star spot images can be detected. By taking the maximum neighborhood value with the gray level of the current scanning pixel as qxq as the judgment condition for starting windowing detection, most of the image processing processes can be filtered under the condition that the real star spot image spot is not leaked to be detected, the image processing calculated amount is reduced, and the detection efficiency is improved. After the detection window is windowed, the self-adaptive threshold of the image is calculated by the part within the neighborhood of n multiplied by n at the center of the detection window, so that an accurate local threshold can be obtained, and the conditions of missing detection and false alarm caused by detection threshold failure due to uneven background gray distribution of the whole frame of image are avoided. According to the standard of detecting the star spot, the isolated noise in the background image can be accurately screened out, and the star spot meeting the requirements is detected. According to the invention, threshold calculation is carried out according to the current scanning pixel neighborhood, more accurate segmentation threshold can be obtained, the used star spot detection standard logic is simple, and the FPGA logic realization is facilitated.