A cross-silk center positioning method based on interference image fringe background elimination

Abstract

The present application relates to a kind of cross-silk center positioning method based on interference image fringe background elimination, mainly used to solve the technical problem that cross-silk is prone to confusion with fringe background in interference image, leading to difficult to accurately locate cross-silk center coordinate.The present application eliminates the interference of interference image fringe background, and then accurately locates cross-silk center by using edge detection to identify cross-silk.By exploring the potential connection between interference image fringe background and cross-silk imaging, the relative radiation correction method is used to adjust the brightness distribution of interference image, so that the brightness of different interference dimensions tends to be consistent, effectively eliminates the influence of dark fringe background in interference image, avoids confusion with cross-silk image, and avoids inaccurate cross-silk identification, which lays a foundation for subsequent accurate positioning of cross-silk center.

Description

Technical Field

[0001] This invention relates to interferometric image processing methods, specifically to a method for locating the crosshair center based on the elimination of background stripes in interferometric images. Background Technology

[0002] Large-aperture static interferometric imaging is a spatiotemporal joint modulation method for static interferometric imaging. Large-aperture static interferometric spectrometers developed based on this technology exhibit high stability and high sensitivity. This technique achieves hyperspectral data imaging through the transformation of target radiation to interferometric data and the transformation of interferometric data to spectral data. Therefore, the quality of the interferometric data is crucial to the final generated hyperspectral data.

[0003] Crosshair center positioning in interferometric imaging aims to improve the precision of z-pointing adjustment during the setup and adjustment of large-aperture static interferometric spectrometers, thereby enhancing the observation capabilities and imaging quality of on-orbit satellites and maintaining the stability and reliability of large-aperture static interferometric spectrometers for long-term onboard use. Common crosshair center positioning methods use a line connecting two points; however, due to the inherent error from manual intervention, the positioning accuracy is low, and batch operations are not feasible, resulting in low efficiency in practical applications.

[0004] In addition, the imaging center point at zero optical path difference in the interferometric image of a large-aperture static interferometric spectrometer theoretically coincides with the imaging position of the crosshairs. However, during the assembly and adjustment process, the crosshairs in the interferometric image are easily confused with the fringe background, making it difficult to accurately locate the coordinates of the crosshair center. Summary of the Invention

[0005] The purpose of this invention is to solve the technical problem that crosshairs in interference images are easily confused with the striped background, making it difficult to accurately locate the center coordinates of the crosshairs. A method for locating the center of the crosshairs based on the elimination of striped background in interference images is proposed.

[0006] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows:

[0007] A method for locating the crosshair center based on interference image fringe background elimination, characterized by the following steps:

[0008] S1, Read the M frames of the original interference image and perform averaging along the frame dimensions to obtain a stable interference image containing the crosshairs, as shown in the following formula:

[0009]

[0010] Among them, F n Let F represent the original interference image of the nth frame;

[0011] S2, Obtain the crosshair image I with striped background removed. ′ :

[0012] S2.1, Select region of interest I in the interferometric image, and calculate the mean vector m = (m1, m2, ..., m) along the width dimension of region of interest I to obtain the mean vector of different fringe backgrounds. i ,…,m h ) T m i Characterizes the brightness level of the i-th row of stripes in region of interest I:

[0013]

[0014] Where h represents the number of rows in region of interest I, w represents the number of columns in region of interest I, and p (i,j) This represents the pixel value in the i-th row and j-th column of the region of interest I;

[0015] S2.2, average the mean vector m to obtain the global mean m. all The overall brightness level of region of interest I is determined using the following formula:

[0016]

[0017] S2.3, the fringe background of region of interest I is eliminated using the relative radiometric correction method to obtain the crosshair image I. ′ The formula is as follows:

[0018]

[0019] S3, for crosshair image I ′ Edge detection is performed on the center of the crosshairs to obtain a binary edge map BW. The set of straight lines L that constitute the crosshairs is extracted from the binary edge map BW. Based on the set of straight lines L, the centroid coordinates and centroid coordinates of the center of the crosshairs are accurately located, thus completing the positioning of the center of the crosshairs.

[0020] Furthermore, step S3 specifically includes:

[0021] S3.1, the Canny algorithm is used to process the crosshair image I. ′ Edge detection is performed to obtain the binary edge map BW, as shown in the following formula:

[0022] BW = Canny(I) ′ );

[0023] S3.2, using the Hough transform, extract the set of lines L that form the crosshairs from the binary edge graph BW, as shown in the following formula:

[0024] L = Hough(BW);

[0025] S3.3, Calculate the centroid c of the crosshairs based on the extracted set of lines L. (x,y) and center of mass b (x,y) The formula is as follows:

[0026]

[0027] Where f(L) represents the function for calculating the centroid of the crosshairs, N represents the number of intersecting pixels of the crosshairs, g(L) represents the function for calculating the centroid of the crosshairs, H represents the height of the region enclosed by the N intersecting pixels of the crosshairs, and W represents the width of the region enclosed by the N intersecting pixels of the crosshairs.

[0028] Further, in step S2.1, the region of interest I is selected in the interferometric image, specifically as follows:

[0029] In the interferometric image, a sub-region containing the crosshairs is selected as the region of interest (ROI), and the coordinate offset of ROI is recorded. The coordinate offset is the upper left corner offset of ROI. The upper left corner offset is used as the origin of the coordinate system of ROI, as shown in the following formula:

[0030] I = roi(F);

[0031] Here, roi() is the location function for the region of interest.

[0032] Furthermore, it also includes step S4:

[0033] S4, based on the coordinate offset of the region of interest I, reposition the centroid and centroid positions calculated in step S3.3 to the interferometric image, respectively c ′ (x,y) and b ′ (x,y) The formula is as follows:

[0034] c ′ (x,y) =c (x,y) +offset;

[0035] b ′ (x,y) =b (x,y) +offset.

[0036] Further, in step S2.1, a sub-region centered on the crosshairs is selected as the region of interest I in the interference image.

[0037] Furthermore, in step S3.1, the Canny algorithm is used to process the crosshair image F. ′ Perform single-line double-edge detection. In step 3.3, N=4.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1. This invention explores the potential connection between the fringe background of the interference image and the crosshair imaging. It uses the relative radiometric correction method to adjust the brightness distribution of the interference image, so that the brightness of different interference dimensions tends to be consistent, effectively eliminating the influence of the dark fringe background in the interference image, avoiding confusion with the crosshair image and causing inaccurate crosshair identification, and laying a solid foundation for the subsequent accurate positioning of the crosshair center.

[0040] 2. In this invention, the crosshair image I ′ When performing edge detection, the crosshair recognition of the extended interferometric image is transformed into single-line double-edge detection, which further improves the accuracy of crosshair center positioning in the interferometric image.

[0041] 3. By selecting the region of interest in the interferometric image, this invention allows us to focus only on a specific part of the interferometric image, enabling targeted analysis and calculation of the region of interest, avoiding redundant calculations, and improving computational efficiency. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating an embodiment of a crosshair center localization method based on interference image stripe background elimination according to the present invention;

[0043] Figure 2 This is a schematic diagram of frame-by-frame reading and averaging of an interference image in an embodiment of a crosshair center localization method based on interference image stripe background elimination according to the present invention;

[0044] Figure 3 This is a schematic diagram illustrating the removal of striped background in the region of interest in an embodiment of a crosshair center localization method for removing striped background based on interference image stripe background according to the present invention;

[0045] Figure 4 This is a flowchart illustrating the process of locating the crosshair center in an interference image, as described in an embodiment of the present invention, which is a method for locating the crosshair center in an interference image based on the elimination of background stripes in an interference image. Detailed Implementation

[0046] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0047] This invention proposes a method for locating the crosshair center based on interference image fringe background elimination. It explores the potential connection between the interference image fringe background and crosshair imaging, and eliminates the interference of dark fringe background by adjusting the image brightness distribution. Figure 1 As shown, the specific steps include:

[0048] S1, Read the M frames of the original interference image and average them along the frame dimensions to reduce the noise of the original interference image and obtain a stable interference image containing crosshairs;

[0049] The data of the raw interferometric image consists of a frame header and data for each frame, such as... Figure 2 As shown, the interference image is obtained by averaging each frame of data using the following formula:

[0050]

[0051] Among them, F n Let F represent the original interference image of the nth frame;

[0052] S2, as Figure 3 As shown, eliminating the striped background in region of interest I specifically includes the following steps:

[0053] S2.1 To improve computational efficiency, a sub-region containing the crosshairs is selected in the interferometric image as the region of interest (ROI), and the coordinate offset of ROI is recorded. The coordinate offset refers to the upper left corner coordinate offset of ROI in the interferometric image; the upper left corner coordinate offset is taken as the origin of the coordinate system of ROI.

[0054] In this embodiment, the sub-region centered on the crosshair is selected as the region of interest I, as shown in the following formula:

[0055] I = roi(F);

[0056] Where roi() is the location function for the region of interest;

[0057] The mean vector m = (m1, m2, ..., m) of different striped backgrounds in region of interest I is obtained by averaging along the width dimension. i ,…,m h ) T m i Characterizes the brightness level of the i-th row of stripes in region of interest I:

[0058]

[0059] Where h represents the number of rows in region of interest I, w represents the number of columns in region of interest I, and p (i,j) This represents the pixel value in the i-th row and j-th column of the region of interest I;

[0060] S2.2, further average the mean vector m to obtain the global mean m of the region of interest I. all The overall brightness level of region of interest I is determined using the following formula:

[0061]

[0062] S2.3, The relative radiometric correction method is used to eliminate the fringe background in region of interest I, resulting in a clear crosshair image I. ′ The formula is as follows:

[0063]

[0064] S3, as Figure 4 As shown, for crosshair image I ′ The center of the crosshairs is precisely located to obtain the centroid and centroid coordinates of the crosshairs; this specifically includes the following steps:

[0065] S3.1, the Canny algorithm is used to process the crosshair image I. ′ Edge detection is performed; in this embodiment, it is set to single-line double-edge detection to obtain the binary edge map BW, as shown in the following formula:

[0066] BW = Canny(I) ′ );

[0067] The Canny algorithm is widely used in computer vision, and it can reduce image noise while maintaining accurate edge localization. The specific steps of the Canny algorithm are as follows:

[0068] (1) Noise filtering is performed using a Gaussian filter;

[0069] (2) The Sobel operator calculates the image gradient to determine the edge direction and intensity;

[0070] (3) Apply non-maximum suppression in the gradient direction to retain the point with the maximum gradient in that direction;

[0071] (4) The dual threshold method distinguishes between strong edges, weak edges, and non-edges;

[0072] (5) Connect strong edges and weak edges to form a complete edge;

[0073] S3.2, using the Hough transform, extract the set of lines L that form the crosshairs from the binary edge graph BW, as shown in the following formula:

[0074] L = Hough(BW);

[0075] The Hough transform is an image processing technique used to detect geometric shapes (such as lines and circles) in an image. Its basic principle is as follows:

[0076] (1) Convert the pixels in the image into a geometric representation in the parameter space (Hough space); for example, for detecting a line, a point in the parameter space represents a line.

[0077] (2) For edge points in the image, the point is mapped to multiple candidate geometric shapes in the parameter space according to different parameterization descriptions (e.g., line equations), and the votes are accumulated on the parameter values ​​corresponding to these candidate shapes.

[0078] (3) Find the local maximum values ​​of the accumulated votes in the parameter space, which correspond to the geometry of the image.

[0079] S3.3, Calculate the centroid c of the crosshairs based on the extracted set of lines L. (x,y) (i.e., geometric center) and centroid b (x,y) The formula is as follows:

[0080]

[0081]

[0082] Where f(L) represents the function for calculating the centroid of the crosshairs, and N represents the number of pixels where the crosshairs intersect. Since crosshair recognition is a single-line double-edge detection, N = 4.

[0083] g(L) represents the function for calculating the centroid of the crosshairs. The centroid position is obtained by weighted calculation of the region enclosed by the above four intersection points. H represents the height of the enclosed region, and W represents the width of the enclosed region.

[0084] S4, based on the coordinate offset of the region of interest I, reposition the centroid and centroid positions calculated in step S3.3 to the interferometric image, respectively c ′ (x,y) and b ′ (x,y) The formula is as follows:

[0085] c ′ (x,y) =c (x,y) +offset;

[0086] b ′ (x,y) =b (x,y) +offset.

Claims

1. A method for locating the center of a crosshair based on interference image fringe background elimination, characterized in that, Includes the following steps: S1, Read the M frames of the original interference image and perform averaging along the frame dimensions to obtain a stable interference image containing the crosshairs, as shown in the following formula: Among them, F n Let F represent the original interference image of the nth frame; S2, Obtain the crosshair image I′ with the striped background removed: S2.1, Select region of interest I in the interferometric image, and calculate the mean vector m = (m1, m2, ..., m) along the width dimension of region of interest I to obtain the mean vector of different fringe backgrounds. i ,…,m h ) T m i Characterizes the brightness level of the i-th row of stripes in region of interest I: Where h represents the number of rows in region of interest I, w represents the number of columns in region of interest I, and p (i,j) This represents the pixel value in the i-th row and j-th column of the region of interest I; S2.2, average the mean vector m to obtain the global mean m. all The overall brightness level of region of interest I is determined using the following formula: S2.3, the fringe background of the region of interest I is eliminated using the relative radiometric correction method to obtain the crosshair image I′, as shown in the following formula: S3. Perform edge detection on the center of the crosshair in the crosshair image I′ to obtain the edge binary map BW. Extract the set of straight lines L that constitute the crosshair from the edge binary map BW. Accurately locate the center of the crosshair based on the set of straight lines L to obtain the centroid coordinates and centroid coordinates of the center of the crosshair, thus completing the localization of the center of the crosshair.

2. The method for crosshair center localization based on interference image fringe background elimination according to claim 1, characterized in that, Step S3 specifically includes: S3.1, the Canny algorithm is used to perform edge detection on the crosshair image I′ to obtain the binary edge map BW, as shown in the following formula: BW = Canny(I'); S3.2, using the Hough transform, extract the set of lines L that form the crosshairs from the binary edge graph BW, as shown in the following formula: L = Hough(BW); S3.3, Calculate the centroid c of the crosshairs based on the extracted set of lines L. (x,y) and center of mass b (x,y) The formula is as follows: Where f(L) represents the function for calculating the centroid of the crosshairs, N represents the number of intersecting pixels of the crosshairs, g(L) represents the function for calculating the centroid of the crosshairs, H represents the height of the region enclosed by the N intersecting pixels of the crosshairs, and W represents the width of the region enclosed by the N intersecting pixels of the crosshairs.

3. The method for crosshair center localization based on interference image fringe background elimination according to claim 1, characterized in that, In step S2.1, the region of interest I is selected in the interferometric image, specifically as follows: In the interferometric image, a sub-region containing the crosshairs is selected as the region of interest (ROI), and the coordinate offset of ROI is recorded. The coordinate offset is the upper left corner offset of ROI. The upper left corner offset is used as the origin of the coordinate system of ROI, as shown in the following formula: I = roi(F); Here, roi() is the location function for the region of interest.

4. The method for crosshair center localization based on interference image fringe background elimination according to claim 3, characterized in that, It also includes step S4: S4, based on the coordinate offset of the region of interest I, reposition the centroid and mass center positions calculated in step S3.3 to the interferometric image, respectively, as c′ (x,y) and b′ (x,y) The formula is as follows: c′ (x,y) =c (x，y) +offset; b′ (x，y) =b (x，y) +offset。 5. The method for crosshair center localization based on interference image fringe background elimination according to claim 4, characterized in that: In step S2.1, a sub-region centered on the crosshairs is selected as the region of interest I in the interference image.

6. The method for crosshair center localization based on interference image fringe background elimination according to claim 5, characterized in that: In step S3.1, the Canny algorithm is used to perform single-line double-edge detection on the crosshair image F′. In step 3.3, N = 4.

Images