A method for locating the centroid of space targets

Abstract

The present invention provides a method for locating the centroid of a space target, comprising: step 1, obtaining the circumscribed rectangle of the space target in the image, and segmenting the space target from the background area; step 2, performing edge detection on the space target by using a gradient operator, fitting the edge points into a straight line to obtain the sailboard part of the space target; step 3, pairing the edge points, fitting the distance of the paired edge points into a straight line, obtaining the boundary line between the space target body and the sailboard, separating the space target body from the sailboard through the boundary line, and obtaining the space target body; step 4, completing the acquisition of the minimum circumscribed rectangle of the local space target based on the principal axis method, calculating the centroid of the minimum circumscribed rectangle, and completing the centroid positioning of the space target. The present invention can quickly obtain the minimum circumscribed rectangle of the space target, with an error within two pixels, thereby completing the centroid positioning of the space target.

Description

Technical Field

[0001] The invention relates to the field of digital image processing, and in particular to a method for locating the centroid of a space target. Background Art

[0002] Space target monitoring is of great significance for maintaining limited orbital resources and avoiding space collisions. As one of the means of monitoring, space-based photoelectric detection is widely used in the field of space target monitoring due to its flexible and maneuverable characteristics. Detecting space targets through photoelectric observation and imaging of space-based systems is an important means of space target monitoring. However, there are few detection and positioning methods for space targets in the current target detection field, and their applicability is not strong. Traditional image processing algorithms and methods based on deep learning cannot achieve good detection results.

[0003] There are currently two detection methods for space targets, one is a method based on grayscale information, and the other is a method based on regional fitting. The positioning method based on grayscale information is based on the grayscale information of the target, and then obtains the center of mass position of the target. There are mainly grayscale center of mass method, grayscale center of mass method with threshold, square weighted center of mass method, etc. Since the positioning accuracy of this type of algorithm is related to the grayscale information of the target, this method is suitable for targets with relatively uniform grayscale distribution. Because the grayscale of the space target is related to the material properties and will show different brightness at different angles, the grayscale feature is not stable in the time domain, and is generally based on regional fitting methods. The existing algorithm considers space targets that are imaged as circles or ellipses. The positioning methods include Gaussian surface fitting method, improved Gaussian surface fitting method, circle (ellipse) fitting, and fitting method based on geometric features. Compared with long-distance detection targets, in close-range detection, the center of mass of the target can be more accurately located through rich information, and the shape, area, color and other information of the target can be obtained. With the improvement of the performance of optoelectronic telescopes, it is now possible to capture high-resolution images of space targets, based on which space targets can be located, their status analyzed, and their three-dimensional posture estimated, which is of great significance in space situational awareness. Summary of the invention

[0004] In view of the problems existing in the prior art, a method for locating the centroid of a space target is provided, which completes the centroid positioning for a space target with more specific imaging.

[0005] The technical solution adopted by the present invention is as follows: A method for locating the centroid of a space target, comprising:

[0006] Step 1: Obtain the bounding rectangle of the space target in the image and separate the space target from the background area;

[0007] Step 2: Use a gradient operator to detect the edge of the space target, and fit the edge points into a straight line to obtain the sailboard part of the space target;

[0008] Step 3: Pair the edge points, fit the distance of the paired edge points into a straight line, obtain the boundary line between the space target body and the sailboard, separate the space target body from the sailboard through the boundary line, and obtain the space target body;

[0009] Step 4: Based on the principal axis method, the local minimum bounding rectangle of the space target is obtained, the centroid of the minimum bounding rectangle is calculated, and the centroid positioning of the space target is completed.

[0010] Further, the sub-steps of step 1 include:

[0011] Step 1.1, binarize the image;

[0012] Step 1.2, performing expansion operation on the spatial object in the image to obtain an expanded image;

[0013] Step 1.3, perform erosion operation on the dilated image to obtain an eroded image, and traverse the eroded image to obtain the coordinates of the lower right corner of the spatial target;

[0014] Step 1.4, traverse the expansion image according to the coordinates of the lower right corner of the space target to obtain the coordinates of the upper left corner of the space target;

[0015] Step 1.5: According to the coordinates of the upper left corner and the lower right corner of the space target, the circumscribed rectangle of the space target can be obtained.

[0016] Furthermore, in step 1.3, the spatial target in the image is reduced to an isolated point of 1X1 through corrosion operation, the corroded image is traversed, and the point with a value of 1 is detected as the coordinate of the lower right corner of the spatial target.

[0017] Furthermore, the specific steps of step 1.4 are:

[0018] Step 1.4.1, mark the coordinate of the lower right corner as a, start from coordinate a and search upwards, stop after finding the last point with a pixel value of 1, and get coordinate b;

[0019] Step 1.4.1, search from coordinate b to the left, stop after finding the last point with a pixel value of 1, and get coordinate c;

[0020] Step 1.4.1, search upward from coordinate c, take the column coordinate of coordinate c as the starting column, take the column coordinate of coordinate a as the ending column, count whether there is a point with a pixel value of 1 from the starting column to the ending column, if there is, continue to search upward, otherwise stop, and get coordinate d;

[0021] Step 1.4.1, search from coordinate d to the left, with the row coordinate of coordinate d as the starting row and the row coordinate of coordinate a as the ending row. Count whether there is a point with a pixel value of 1 from the starting to the ending rows. If there is, continue to search to the left, otherwise stop and get the coordinate e, which is the coordinate of the upper left corner of the space target.

[0022] Furthermore, the specific process of step 2 is as follows:

[0023] Step 2.1, using the image gradient operator to detect the edge of the target;

[0024] Step 2.2: Use the RANSAC straight line fitting method to fit the edge points into two straight lines, namely, the straight lines describing the sailboard of the space target.

[0025] Furthermore, the specific sub-steps of step 3 are:

[0026] Step 3.1, pair edge points along the perpendicular line of the straight line of the sailboard, and calculate the Euclidean distance of each pair of edge points;

[0027] Step 3.2: Draw a scatter plot of the Euclidean distance of each pair of edge points in ascending order of the horizontal coordinates and fit them into a straight line ;

[0028] Step 3.3, through the straight line The five parts of the space target can be distinguished, which are the transition section of the left sailboard, the stable section of the left sailboard, the main body section, the stable section of the right sailboard, and the transition section of the right sailboard, thereby obtaining the boundary line between the main body of the space target and the sailboard;

[0029] Step 3.3: Separate the space target along the dividing line to obtain the space target body.

[0030] Furthermore, in step 3.1, the vertical angle of the straight line of the sailboard is used as the scanning direction, and each pair of edge points is located on the same scanning line; the Euclidean distance calculation method for each pair of edge points is:

[0031]

[0032] in, are the horizontal and vertical coordinates of the i-th edge point, are the coordinates of the j-th edge point respectively, and the ith and j-th edge points are paired edge points.

[0033] Furthermore, the sub-steps of step 4 are:

[0034] Step 4.1, determine the horizontal principal axis, vertical principal axis and rotation center of the space target body;

[0035] Step 4.2, calculate the circumscribed rectangle of the space target body;

[0036] Step 4.3, rotate the vertical main axis counterclockwise with the horizontal main axis as the rotation axis, and the rotation angle is 1° each time. Repeat steps 4.1 to 4.2 until one circle is rotated to obtain the circumscribed rectangle of the space target body at different angles;

[0037] Step 4.4, compare the circumscribed rectangles of the space object body at different angles to obtain the minimum circumscribed rectangle;

[0038] Step 4.5, calculate the centroid of the minimum circumscribed rectangle, that is, the centroid of the space target, and complete the centroid positioning of the space target.

[0039] Furthermore, in step 4.1, the space target is regarded as a flat plate with uniform density, and the centroid of each row is fitted into a straight line as the vertical axis; the centroid of each column is fitted into a straight line as the horizontal axis, and the intersection of the horizontal axis and the vertical axis is the rotation center.

[0040] Furthermore, in step 4.2, the calculation method of the circumscribed rectangle is as follows: keep the direction of the horizontal main axis unchanged, translate the horizontal main axis upward, and stop when there is no intersection between the horizontal main axis and the space target body, and use it as the upper boundary of the circumscribed rectangle; move the horizontal main axis downward, and stop when there is no intersection between the horizontal main axis and the space target body, and use it as the lower boundary of the circumscribed rectangle; use the vertical line of the horizontal main axis as the slope, draw a straight line through the rotation center, translate the straight line to the left, and stop when there is no intersection between the straight line and the space target body, and use it as the left boundary of the circumscribed rectangle; translate the straight line to the right, and stop when there is no intersection between the straight line and the space target body, and use it as the right boundary of the circumscribed rectangle, and the circumscribed rectangle can be obtained.

[0041] Compared with the prior art, the beneficial effects of adopting the above technical solution are: the present invention does not require buffer time, can quickly obtain the minimum bounding rectangle of the space target, and the error is within two pixels, thereby completing the centroid positioning of the space target. It does not require the support of the image processing library, has low hardware requirements, and is applicable to a variety of hardware platforms. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A flow chart of a method for locating the centroid of a space target is proposed for the present invention.

[0043] Figure 2 It is a structural element expanded in one embodiment of the present invention.

[0044] Figure 3 It is a template covered by the expanded image in one embodiment of the present invention.

[0045] Figure 4 It is a structural element corroded in one embodiment of the present invention.

[0046] Figure 5 It is a template covered by the image after corrosion in one embodiment of the present invention.

[0047] Figure 6 Schematic diagram of upper left corner coordinate calculation in one embodiment of the present invention.

[0048] Figure 7 FIG. 4 is a schematic diagram of a circumscribed rectangle in an embodiment of the present invention.

[0049] Figure 8 Schematic diagram of a gradient operator in one embodiment of the present invention.

[0050] Fig. 9 Schematic diagram of straight line detection of a windsurfing board in one embodiment of the present invention.

[0051] Fig.10 FIG. 1 is a schematic diagram of pairing edge points in an embodiment of the present invention.

[0052] Fig.11 It is a schematic diagram of a statistical diagram of the distance between paired edge points and a straight line fitting result in one embodiment of the present invention.

[0053] Fig.12 FIG. 4 is a segmented schematic diagram of the distance between paired points in one embodiment of the present invention.

[0054] Fig.13 It is a schematic diagram of extracting the main body of a space target in one embodiment of the present invention.

[0055] Fig.14 Schematic diagram of the minimum circumscribed rectangle detection result of the space target body in one embodiment of the present invention.

[0056] Fig.15 Error statistics for centroid calculations using the present invention. DETAILED DESCRIPTION

[0057] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar modules or modules with the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be construed as limitations on the present application. On the contrary, the embodiments of the present application include all changes, modifications and equivalents that fall within the spirit and connotation of the appended claims.

[0058] When the detector is close to the space target, the target is a face image in the final imaging, and the imaging contains a lot of details of the close-range space target, including contour information, texture information, color information or grayscale information, etc. The high-resolution imaging of the space target can clearly identify the contour of the satellite body and the solar panel part. Therefore, in this embodiment, the centroid position of the space target is detected by using a method based on geometric features. The segmentation of the target body and the solar panel is first achieved through coarse recognition, and then the centroid position is detected through the description of the minimum circumscribed rectangle.

[0059] Before extracting the target centroid coordinate information, the target in the image must be roughly identified. Rough identification is actually an image segmentation process, which requires obtaining the approximate position of the target from the image. Rough identification is divided into three steps: separating the space target from the background area and extracting the circumscribed rectangle of the space target; detecting the sailboard to obtain the imaging straight line of the sailboard; separating the target body from the sailboard to obtain the body contour of the space target. After rough identification, the minimum circumscribed rectangle method is used to accurately locate the centroid. The specific plan is as follows:

[0060] like Figure 1 As shown, a method for locating the centroid of a space target includes:

[0061] Step 1: Obtain the bounding rectangle of the space target in the image and segment the space target from the background area; the subsequent steps are performed based on the image within the bounding rectangle;

[0062] Step 2: Use a gradient operator to detect the edge of the space target, and fit the edge points into a straight line to obtain the sailboard part of the space target;

[0063] Step 3: Pair the edge points, fit the distance of the paired edge points into a straight line, obtain the boundary line between the space target body and the sailboard, separate the space target body from the sailboard through the boundary line, and obtain the space target body;

[0064] Step 4: Based on the principal axis method, the local minimum bounding rectangle of the space target is obtained, the centroid of the minimum bounding rectangle is calculated, and the centroid positioning of the space target is completed.

[0065] Since high-resolution images provide more detailed information about space targets, they also bring more image calculations; at the same time, high-frame-rate image sequences place higher demands on the real-time performance of the algorithm. In order to accurately locate the target center of mass without redundant calculations, the image area that needs to be processed can be reduced by extracting the circumscribed rectangle of the space target. Specifically, step 1 includes the following sub-steps:

[0066] Step 1.1, binarize the image;

[0067] Step 1.2, performing expansion operation on the spatial object in the image to obtain an expanded image;

[0068] The expansion operation is used to expand the interior of the space target. In this embodiment, the expanded structural element takes into account the image details in different directions. The specific form is as follows: Figure 2 As shown in , there are six kinds of structural elements. Each expansion structural element is a 2*2 matrix. Four of them have pixel values ​​of 0 in the upper left corner, lower left corner, upper right corner, and lower right corner, and pixel values ​​of other positions are 1. The remaining two are unit diagonal matrices and unit anti-diagonal matrices, that is, the pixels of the diagonals are 0 or 1 respectively. Copy an image A that is exactly the same as the image to be expanded. If any one of the structural elements is completely contained in image A, then the position is Figure 3 The expansion template shown covers; if it is not included, the pixel value at that position remains unchanged, where the expansion template is a 2*2 matrix with all pixel values ​​1.

[0069] Step 1.3, perform erosion operation on the dilated image to obtain a eroded image, and traverse the eroded image to obtain the coordinates of the lower right corner of the spatial target;

[0070] The space target is reduced to an isolated point of 1×1 by using an erosion operation to obtain the coordinates of the lower right corner of the space target. In the embodiment, the eroded structural element is selected as follows: Figure 4 As shown in the figure, it is a 3*3 matrix, where only the center pixel value is 1, and the rest of the pixel values ​​are 0. Copy an image B that is exactly the same as the image to be eroded (i.e. the image after the previous step of expansion). If the structural element is inconsistent with image B, the position is Figure 5 The corrosion template is shown as shown, where the corrosion template is a 3*3 matrix with all pixel values ​​​​being 0; if it is consistent (that is, it means that the point is a 1×1 isolated point), the pixel value at this position remains unchanged. Traverse the entire image B, and detect the point with a value of 1 to obtain the coordinates of the lower right corner of the spatial target.

[0071] Step 1.4, traverse the expansion image according to the coordinates of the lower right corner of the space target to obtain the coordinates of the upper left corner of the space target;

[0072] Step 1.5: According to the coordinates of the upper left corner and the lower right corner of the space target, the circumscribed rectangle of the space target can be obtained, such as Figure 7 shown.

[0073] The spatial target and background in the image are segmented by the bounding rectangle, and subsequent processing is based on the image area within the bounding rectangle to reduce the amount of calculation.

[0074] Since the lower right corner coordinates of the star point are obtained after the corrosion operation, and the area of ​​the star point after expansion shows connectivity, the upper right corner coordinates of the star point can be calculated based on the corrosion image and the lower right corner coordinates of the star point, such as Figure 6 As shown, the specific steps of step 1.4 are:

[0075] Step 1.4.1, mark the coordinates of the lower right corner as , start from coordinate a and search upwards, stop after finding the last point with a pixel value of 1, and get the coordinates ;

[0076] Step 1.4.1, search from coordinate b to the left, stop after finding the last point with a pixel value of 1, and get the coordinates ;

[0077] Step 1.4.1, search upward from coordinate c, take the column coordinate of coordinate c as the starting column, take the column coordinate of coordinate a as the ending column, count whether there is a point with a pixel value of 1 from the starting column to the ending column, if there is, continue to search upward, otherwise stop, and get the coordinate ;

[0078] Step 1.4.1, search from coordinate d to the left, use the row coordinate of coordinate d as the starting row, and the row coordinate of coordinate a as the ending row, and count whether there is a point with a pixel value of 1 from the starting to the ending row. If there is, continue to search to the left, otherwise stop and get the coordinates , that is, the coordinate of the upper left corner of the space target.

[0079] The morphological characteristics of the space target determine that the area of ​​the sailboards on both sides accounts for a large proportion, which will affect the precise centroid positioning. In order to ensure that the center of mass coordinates are not affected by the sailboard part, it is necessary to locate the position of the sailboard in the image and remove it from the image. Therefore, it is necessary to first identify the two parts of the solar panel and the main body, that is, it is necessary to separate the target body from the sailboards on both sides. Since the sailboards on both sides occupy most of the space target and have obvious straight line features, the sailboard part can be extracted by extracting straight lines. Specifically, the specific process of step 2 is:

[0080] Step 2.1: Use an image gradient operator to detect the edge of the target. In this embodiment, the image gradient operator is as follows: Figure 8 As shown, there are commonly used horizontal gradient operators and vertical gradient operators;

[0081] Step 2.2: Use the RANSAC straight line fitting method to fit the edge points into two straight lines, i.e., the straight lines describing the sailboard of the space target, such as Fig. 9 shown.

[0082] Then, the edge points can be combined to complete the extraction of the space target body. The specific sub-steps of step 3 are:

[0083] Step 3.1, pair edge points along the perpendicular line of the straight line of the sailboard, and calculate the Euclidean distance of each pair of edge points;

[0084] In this embodiment, the perpendicular line of the sailboard is used as the scanning direction; the pairing points are defined as edge points on the same scanning line, such as Fig.10 As shown in the figure, P1, P2, P3, and P4 are all edge points on the image. P1 and P2 are on a scanning line, so {P1, P2} is called a pair of paired points; P3 and P4 are on a scanning line, so {P3, P4} is called a pair of paired points. The Euclidean distance between paired points {P1, P2} is obtained according to the following formula :

[0085]

[0086] in, are the coordinates of P1, is the coordinate of P2. Similarly, the distances of all paired points can be calculated. .

[0087] Step 3.2: In order from small to large horizontal coordinates, calculate the Euclidean distance of each pair of edge points. Make a scatter plot and fit it into a straight line through RANSAC ,like Fig.11 As shown;

[0088] Step 3.3, straight line The space target is divided into five sections, namely the transition section of the left sailboard, the stable section of the left sailboard, the main section, the stable section of the right sailboard, and the transition section of the right sailboard. Fig.12 As shown, the scattered points on the straight line are the second and fourth segments corresponding to the stable segments of the left sailboard and the stable segments of the right sailboard, the left side of the second segment is the transition segment of the first segment corresponding to the left sailboard, the right side of the second segment is the third segment corresponding to the main body segment, and the right side of the fourth segment is the transition segment of the fifth segment corresponding to the right sailboard, so the boundary line between the main body and the sailboard of the space target can be obtained;

[0089] Step 3.3: Separate the space target along the dividing line to obtain the space target body. The final body image is as follows: Fig.13 shown.

[0090] After the space target body is separated from the sailboard, the space target body can be positioned. The specific steps of step 4 are:

[0091] Step 4.1, determine the horizontal principal axis, vertical principal axis and rotation center of the space target body;

[0092] In this embodiment, the space target is regarded as a flat plate with uniform density, and the centroid of each row is fitted into a straight line as the vertical axis; the centroid of each column is fitted into a straight line as the horizontal axis, and the intersection of the horizontal axis and the vertical axis is the rotation center.

[0093] Step 4.2, calculate the circumscribed rectangle of the space target body;

[0094] In order to find the upper boundary of the minimum circumscribed rectangle, keep the direction of the horizontal principal axis unchanged, translate the horizontal principal axis upward until it stops when there is no intersection between the horizontal principal axis and the space target body, and use it as the upper boundary of the circumscribed rectangle; move the horizontal principal axis downward until it stops when there is no intersection between the horizontal principal axis and the space target body, and use it as the lower boundary of the circumscribed rectangle; use the vertical line of the horizontal principal axis as the slope, draw a straight line passing through the rotation center, translate this straight line to the left, and stop when there is no intersection between this straight line and the space target body, and use it as the left boundary of the circumscribed rectangle; translate this straight line to the right, and stop when there is no intersection between this straight line and the space target body, and use it as the right boundary of the circumscribed rectangle.

[0095] Step 4.3, rotate the vertical main axis counterclockwise with the horizontal main axis as the rotation axis, and the rotation angle is 1° each time. Repeat steps 4.1 to 4.2 until one circle is rotated to obtain the circumscribed rectangle of the space target body at different angles;

[0096] Step 4.4: Compare the bounding rectangles of the space object at different angles to obtain the minimum bounding rectangle, such as Fig.14 As shown;

[0097] Step 4.5, calculate the centroid of the minimum circumscribed rectangle, that is, the centroid of the space target, and complete the centroid positioning of the space target.

[0098] Finally, the effectiveness of the solution of this embodiment is analyzed through experiments. The experiments prove that the minimum bounding rectangle of the target body can be obtained by using the principal axis method with an error within 2 pixels. The specific error is as follows: Fig.15 As shown in the figure, the calculation time for a single frame image of 5120*5120*8BIT does not exceed 200ms, which meets the real-time requirements.

[0099] It should be noted that in the description of the embodiments of the present invention, unless otherwise clearly specified and limited, the terms "setting" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances; the drawings in the embodiments are used to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations.

[0100] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for locating the centroid of a space target, characterized in that: include: Step 1: Obtain the bounding rectangle of the space target in the image and separate the space target from the background area; Step 2: Use a gradient operator to detect the edge of the space target, and fit the edge points into a straight line to obtain the sailboard part of the space target; Step 3: Pair the edge points, fit the distance of the paired edge points into a straight line, obtain the boundary line between the space target body and the sailboard, separate the space target body from the sailboard through the boundary line, and obtain the space target body; Step 4: Based on the principal axis method, the minimum bounding rectangle of the local space target is obtained, the centroid of the minimum bounding rectangle is calculated, and the centroid positioning of the space target is completed; The specific sub-steps of step 3 are: Step 3.1, pair edge points along the perpendicular line of the straight line of the sailboard, and calculate the Euclidean distance of each pair of edge points; Step 3.2: Draw a scatter plot of the Euclidean distance of each pair of edge points in ascending order of the horizontal coordinates and fit them into a straight line ; Step 3.3, through the straight line The five parts of the space target can be distinguished, which are the transition section of the left sailboard, the stable section of the left sailboard, the main body section, the stable section of the right sailboard, and the transition section of the right sailboard, thereby obtaining the boundary line between the main body of the space target and the sailboard; Step 3.3, separating the space target along the dividing line to obtain the space target body; The sub-steps of step 4 are: Step 4.1, determine the horizontal principal axis, vertical principal axis and rotation center of the space target body; Step 4.2, calculate the circumscribed rectangle of the space target body; Step 4.3, rotate the vertical main axis counterclockwise with the horizontal main axis as the rotation axis, and the rotation angle is 1° each time. Repeat steps 4.1 to 4.2 until one circle is rotated to obtain the circumscribed rectangle of the space target body at different angles; Step 4.4, compare the circumscribed rectangles of the space object body at different angles to obtain the minimum circumscribed rectangle; Step 4.5, calculate the centroid of the minimum circumscribed rectangle, that is, the centroid of the space target, and complete the centroid positioning of the space target.

2. The method for locating the centroid of a space target according to claim 1, characterized in that: The sub-steps of step 1 include: Step 1.1, binarize the image; Step 1.2, performing expansion operation on the spatial object in the image to obtain an expanded image; Step 1.3, perform erosion operation on the dilated image to obtain an eroded image, and traverse the eroded image to obtain the coordinates of the lower right corner of the spatial target; Step 1.4, traverse the expansion image according to the coordinates of the lower right corner of the space target to obtain the coordinates of the upper left corner of the space target; Step 1.5: According to the coordinates of the upper left corner and the lower right corner of the space target, the circumscribed rectangle of the space target can be obtained.

3. The method for locating the centroid of a space target according to claim 2, characterized in that: In the step 1.3, the spatial target in the image is reduced to an isolated point of 1X1 through corrosion operation, the corrosion image is traversed, and the point with a value of 1 is detected as the coordinate of the lower right corner of the spatial target.

4. The method for locating the centroid of a space target according to claim 2 or 3, characterized in that: The specific steps of step 1.4 are: Step 1.4.1, mark the coordinate of the lower right corner as a, start from coordinate a and search upwards, stop after finding the last point with a pixel value of 1, and get coordinate b; Step 1.4.1, search from coordinate b to the left, stop after finding the last point with a pixel value of 1, and get coordinate c; Step 1.4.1, search upward from coordinate c, take the column coordinate of coordinate c as the starting column, take the column coordinate of coordinate a as the ending column, count whether there is a point with a pixel value of 1 from the starting column to the ending column, if there is, continue to search upward, otherwise stop, and get coordinate d; Step 1.4.1, search from coordinate d to the left, with the row coordinate of coordinate d as the starting row and the row coordinate of coordinate a as the ending row. Count whether there is a point with a pixel value of 1 from the starting to the ending rows. If there is, continue to search to the left, otherwise stop and get the coordinate e, which is the coordinate of the upper left corner of the space target.

5. The method for locating the centroid of a space target according to claim 1, characterized in that: The specific process of step 2 is: Step 2.1, using the image gradient operator to detect the edge of the target; Step 2.2: Use the RANSAC straight line fitting method to fit the edge points into two straight lines, namely, the straight lines describing the sailboard of the space target.

6. The method for locating the centroid of a space target according to claim 1, characterized in that: In step 3.1, the vertical angle of the straight line of the sailboard is used as the scanning direction, and each pair of edge points is located on the same scanning line; the Euclidean distance calculation method for each pair of edge points is: in, are the horizontal and vertical coordinates of the i-th edge point, are the coordinates of the j-th edge point respectively, and the ith and j-th edge points are paired edge points.

7. The method for locating the centroid of a space target according to claim 1, characterized in that: In step 4.1, the space target is regarded as a flat plate with uniform density, and the centroid of each row is fitted into a straight line as the vertical axis; the centroid of each column is fitted into a straight line as the horizontal axis, and the intersection of the horizontal axis and the vertical axis is the rotation center.

8. The method for locating the centroid of a space target according to claim 1 or 7, characterized in that: In step 4.2, the calculation method of the circumscribed rectangle is as follows: keep the direction of the horizontal main axis unchanged, translate the horizontal main axis upward until it stops when there is no intersection between the horizontal main axis and the space target body, and use it as the upper boundary of the circumscribed rectangle; move the horizontal main axis downward until it stops when there is no intersection between the horizontal main axis and the space target body, and use it as the lower boundary of the circumscribed rectangle; use the vertical line of the horizontal main axis as the slope, draw a straight line through the rotation center, translate the straight line to the left until it stops when there is no intersection between the straight line and the space target body, and use it as the left boundary of the circumscribed rectangle; translate the straight line to the right until it stops when there is no intersection between the straight line and the space target body, and use it as the right boundary of the circumscribed rectangle, and the circumscribed rectangle can be obtained.

Images