Spatial non-cooperative target feature part identification method
By pre-processing and feature extraction of feature parts of spatial non-cooperative targets, feature invariants are calculated to identify the correct feature part contour, which solves the problem that spatial non-cooperative targets are difficult to identify in hyper-close range relative measurements, and achieves high-precision feature part recognition and relative posture measurement.
Patent Information
- Application Number
- CN202411770503.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-04
- Publication Date
- 2025-05-06
AI Technical Summary
The prior art is difficult to effectively identify and measure characteristic parts on non-cooperational targets in space, especially in hyper-close range relative measurements.
By preprocessing the captured images of feature parts of non-cooperation targets in space, extracting lines, constructing outlines, calculating multiple sets of feature invariants of feature triangles, and identifying the correct feature invariants by judging feature invariants.
It realizes high-precision identification of spatial non-cooperative target feature parts and accurate measurement of relative postures, which is suitable for visual detection and recognition of spatial non-cooperative target feature parts with high symmetry.
Smart Images
Figure CN119942144A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of target recognition and relative posture measurement, and in particular to a method for identifying characteristic parts of a spatial non-cooperative target. Background Art
[0002] Space non-cooperative targets refer to targets that have no prior information (or insufficient prior information) and are not equipped with cooperative markers or communication response equipment. Such space non-cooperative targets usually include: satellites that are not equipped with cooperative interfaces, invalid satellites, foreign spacecraft, space debris, etc. In space missions, ultra-short-range capture, rendezvous and docking are often performed on space non-cooperative targets within 20m of the target object.
[0003] In order to achieve close capture or control of spacecraft in non-cooperative space targets, it is necessary to identify the captureable or key feature parts on the non-cooperative space targets and master the high-precision relative position and relative attitude and other degree of freedom information between the two spacecraft. However, there are no markers specifically used for positioning and identification on the current non-cooperative space target spacecraft. Therefore, special structures such as sails, antenna brackets, and apogee engines installed on such spacecraft can be considered as substitute markers.
[0004] The detection and recognition of this type of alternative markers is a major difficulty in the current ultra-short-range relative measurement technology of space non-cooperative targets. Therefore, it is urgent to develop a method for identifying the characteristic parts of space non-cooperative targets to solve the above problem. Summary of the invention
[0005] The purpose of the present invention is to provide a method for identifying characteristic parts of a space non-cooperative target, which can solve the problem that the space non-cooperative target is difficult to detect and identify in ultra-short range relative measurement.
[0006] To achieve the above object, the present invention provides a method for identifying characteristic parts of a spatial non-cooperative target, comprising:
[0007] S1, preprocessing the captured images of the spatial non-cooperative target feature parts and extracting lines in the captured images;
[0008] S2, construct the closed spatial non-cooperative target feature part contour from the extracted image lines;
[0009] S3, calculating multiple groups of feature invariants of multiple feature triangles selected from the contours of the feature parts of the spatial non-cooperative target;
[0010] S4, identifying the correct feature part contour by judging all feature invariants of the spatial non-cooperative target feature parts.
[0011] Optionally, step S1 includes:
[0012] S1.1, performing contrast stretching preprocessing on the captured image of the spatial non-cooperative target feature part;
[0013] S1.2, extracting lines from the captured image of the characteristic part of the spatial non-cooperative target using a straight line detection method, and obtaining parameters characterizing each line.
[0014] Optionally, the straight line detection method is Hough transform or LSD straight line extraction method; and the parameters characterizing each line are slope and intercept.
[0015] Optionally, step S2 comprises:
[0016] S2.1, screening out characteristic straight lines through the positional relationship between the extracted lines, and determining contour straight lines of several spatial non-cooperative target characteristic parts from the characteristic straight lines;
[0017] S2.2, connecting the contour straight lines end to end to obtain the contour of the closed spatial non-cooperative target feature part.
[0018] Optionally, the screening of characteristic straight lines includes: clustering the extracted lines and refitting them by judging the slope and intercept of each line; and judging the positional relationship between the intersections and endpoints of the fitted lines, and eliminating non-characteristic straight lines to obtain characteristic straight lines.
[0019] Optionally, step S3 includes:
[0020] S3.1, selecting any three adjacent contour lines from the contour of the spatial non-cooperative target feature part, and constructing a first characteristic invariant of a first characteristic triangle based on the three contour lines;
[0021] S3.2, construct the second characteristic triangle, the third characteristic triangle, ..., the Nth characteristic triangle in turn for all remaining contour lines, and for each characteristic triangle, construct the second characteristic invariant, the third characteristic invariant, ..., the Nth characteristic invariant in turn.
[0022] Optionally, the first characteristic variable is: [a / b, a / c, b / c, θ1, θ2, θ3];
[0023] Among them, a, b, c are the three characteristic side lengths of the first characteristic triangle, and θ1, θ2, θ3 are the three included angles of the first characteristic triangle.
[0024] Optionally, if any three selected adjacent contour lines cannot form a feature triangle, they are skipped and not processed.
[0025] Optionally, step S4 includes:
[0026] S4.1, by using the difference in angle parameters in each set of feature invariants, all feature invariants of the obtained spatial non-cooperative target feature parts are compared one-to-one with the known feature quantities of the spatial non-cooperative target feature parts;
[0027] S4.2, removing the feature invariants whose ratio of the scaling ratio to the length parameter exceeds a preset threshold, and obtaining the correct feature invariants, that is, obtaining the contour of the correctly extracted and identified feature part.
[0028] Optionally, the scaling ratio is: characteristic side length / actual side length;
[0029] The ratio of the length parameters is: the ratio of two characteristic side lengths in a characteristic invariant / the ratio of two corresponding actual side lengths in the corresponding actual windsurfing board characteristic quantity.
[0030] In summary, compared with the prior art, the present invention has the following beneficial effects:
[0031] 1. The present invention provides a method for identifying characteristic parts of a spatial non-cooperative target, which forms a method for detecting the relative position and posture of a spatial non-cooperative target by capturing and identifying characteristic parts on the spatial non-cooperative target.
[0032] 2. The method for identifying characteristic parts of spatial non-cooperative targets provided by the present invention has the characteristics of high precision and accurate measurement, and is particularly suitable for visual detection and identification of characteristic parts of spatial non-cooperative targets with high symmetry. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 It is a flow chart of the spatial non-cooperative target characteristic part recognition method of the present invention;
[0034] Figure 2 It is a schematic diagram of the non-cooperative target of the present invention;
[0035] Figure 3 It is a schematic diagram of the straight line extraction result of the present invention;
[0036] Figure 4 It is a schematic diagram of the recognition process of the present invention. DETAILED DESCRIPTION
[0037] The following will be combined with the attached Figures 1 to 4 The technical content, structural features, objectives and effects of the present invention are described in detail through preferred embodiments.
[0038] It should be noted that the drawings are in a very simplified form and use non-precise proportions. They are only used to conveniently and clearly assist in explaining the embodiments of the present invention, and are not used to limit the conditions for the implementation of the present invention. Therefore, they have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical content disclosed by the present invention.
[0039] In the description of the present invention, it should be noted that the terms "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.
[0040] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 mechanical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0041] The present invention provides a method for identifying characteristic parts of spatial non-cooperative targets, such as Figure 1 As shown, the spatial non-cooperative target feature part recognition method includes:
[0042] S1, preprocessing the captured images of the spatial non-cooperative target feature parts and extracting lines in the captured images;
[0043] S2, construct the closed spatial non-cooperative target feature part contour from the extracted image lines;
[0044] S3, calculating multiple groups of feature invariants of multiple feature triangles selected from the contours of the feature parts of the spatial non-cooperative target;
[0045] S4, identifying the correct feature part contour by judging all feature invariants of the spatial non-cooperative target feature parts.
[0046] Wherein, in the step S1, the relationship between the captured image and the spatial non-cooperative target feature part can be considered as an affine transformation.
[0047] Wherein, the step S1 comprises the following steps:
[0048] S1.1, performing contrast stretching preprocessing on the captured images of the spatial non-cooperative target feature parts.
[0049] S1.2, extracting lines from the captured images of the characteristic parts of the spatial non-cooperative target by using a line detection method such as Hough transform (Hough transform) or LSD (Line Segment Detector) line extraction method; obtaining parameters characterizing each line, namely the slope and intercept, by the Hough transform or LSD line extraction method.
[0050] It should be noted that in the extraction of the lines, due to the interference of ambient light and the interference of non-feature parts of non-cooperative targets in space, the lines extracted by the straight line detection method still have the phenomenon of mis-extraction, missed extraction, repeated extraction, etc.
[0051] Wherein, the step S2 comprises the following steps:
[0052] S2.1, screening out characteristic straight lines through the positional relationship between the extracted lines, and determining contour straight lines of several spatial non-cooperative target characteristic parts from the characteristic straight lines.
[0053] Specifically, by judging the slope and intercept of each line, the extracted lines are clustered and refitted; and the positional relationship between the intersections and endpoints of the fitted lines is judged, so as to remove non-feature straight lines generated by mis-extraction and repeated extraction, obtain feature straight lines, and determine the contour straight lines of several spatial non-cooperative target feature parts.
[0054] S2.2, connecting the contour straight lines end to end to obtain the contour of the closed spatial non-cooperative target feature part.
[0055] Wherein, the step S3 comprises the following steps:
[0056] S3.1 Select any three adjacent contour lines from the contour of the spatial non-cooperative target feature part, and construct the first characteristic invariant of the first characteristic triangle based on the three contour lines.
[0057] Specifically, using geometric constraint relationships, any three adjacent contour lines are selected to form a first characteristic triangle, the three characteristic side lengths a, b, and c of the first characteristic triangle are measured, and the ratio of any two characteristic side lengths and the angle between any two characteristic sides of the first characteristic triangle are calculated, that is, the first characteristic invariant of the first characteristic triangle: [a / b, a / c, b / c, θ1, θ2, θ3]; wherein θ1, θ2, θ3 are the three angles of the first characteristic triangle.
[0058] Optionally, if any three selected adjacent contour lines cannot form a feature triangle, they are skipped and not processed.
[0059] S3.2, construct the second characteristic triangle, the third characteristic triangle, ..., the Nth characteristic triangle in turn for all remaining contour lines, and for each characteristic triangle, construct the second characteristic invariant, the third characteristic invariant, ..., the Nth characteristic invariant in turn.
[0060] Specifically, any three adjacent contour lines from the remaining contour lines are selected in turn to form a second characteristic triangle, a third characteristic triangle, ..., the Nth characteristic triangle, the three characteristic side lengths of each characteristic triangle are measured, and the ratio of any two characteristic side lengths of the second characteristic triangle and the angle between any two characteristic sides, the ratio of any two characteristic side lengths of the third characteristic triangle and the angle between any two characteristic sides, ..., the ratio of any two characteristic side lengths of the Nth characteristic triangle and the angle between any two characteristic sides are calculated, so as to obtain the second characteristic invariant, the third characteristic invariant, ..., the Nth characteristic invariant.
[0061] Wherein, the step S4 comprises the following steps:
[0062] S4.1, by using the difference in angle parameters in each set of feature invariants, all feature invariants of the obtained spatial non-cooperative target feature parts are compared one-to-one with the known feature quantities of the spatial non-cooperative target feature parts;
[0063] S4.2, remove the feature invariants whose scaling ratio (feature side length / actual side length) and length parameter ratio (ratio of two feature side lengths in a feature invariant / ratio of two actual side lengths corresponding to the corresponding actual sailboard feature quantity) exceed the preset threshold, and obtain the correct feature invariant, that is, obtain the contour of the correctly extracted and identified spatial non-cooperative target feature part.
[0064] like Figure 2 As shown, in a specific embodiment of the present invention, the characteristic part of the space non-cooperative target is a sailboard, and the sailboard is in an octagonal shape.
[0065] The process of identifying the characteristic parts of the windsurfing board is as follows:
[0066] Step S1, contrast stretching is performed on the captured image of the windsurfing board to improve the recognition of the edge line and background of the windsurfing board image; the LSD straight line extraction method is used to perform straight line detection, and the straight line detection result is as follows: Figure 3 As shown; and the slope and intercept characterizing each extracted line are obtained by the LSD straight line extraction method.
[0067] Among them, the green line segments represent the lines extracted from the windsurfing image; due to the interference of ambient light and the interference of non-feature parts of non-cooperative targets in space, there are also line segments of the windsurfing image edge lines that are erroneously extracted, missed, and repeatedly extracted in the green line segments.
[0068] Step S2, by judging the slope and intercept of each extracted line, clustering the extracted lines and refitting them; and judging the positional relationship between the intersection points and endpoints of the fitted lines, thereby removing non-feature straight lines generated by erroneous extraction and repeated extraction, so as to obtain feature straight lines and determine the contour straight line of the windsurfing board;
[0069] The outline of the sailboard is connected end to end to obtain the outline of the closed sailboard, such as Figure 4 shown.
[0070] Step S3, selecting any three adjacent contour lines from the windsurfing board contour, and constructing the first characteristic invariant of the first characteristic triangle based on the three contour lines. Figure 4 The three straight lines 2, 5 and 9 shown constitute the first characteristic triangle in the yellow dotted area. The three characteristic side lengths a, b, c of the first characteristic triangle are measured, and the ratio of any two characteristic side lengths and the angle between any two characteristic sides of the first characteristic triangle are calculated to obtain the corresponding first characteristic invariant: [a / b, a / c, b / c, θ1, θ2, θ3]; in a specific embodiment of the present invention, the first characteristic invariant is [128.467, 126.01, 180.26, 89.81, 44.35, 45.45];
[0071] Among them, a set of actual windsurfing characteristic quantities corresponding to the first characteristic invariant is [a1 / b1, a1 / c1, b1 / c1, θ1 ’ ,θ2 ’ ,θ3 ’ ], the actual side lengths of this set of sailboards are a1, b1, c1.
[0072] According to this method, the second characteristic triangle, the third characteristic triangle, ..., the Nth characteristic triangle are constructed for all the remaining windsurfing contour lines in turn, and the second characteristic invariant, the third characteristic invariant, ..., the Nth characteristic invariant are obtained.
[0073] Step S4, compare the obtained N groups of characteristic invariants with the actual N groups of sailboard characteristic quantities one by one, and make a ratio between the three characteristic side lengths a, b, c of the characteristic triangle corresponding to any group of characteristic invariants and the three actual side lengths a1, b1, c1 in the corresponding group of sailboard characteristic quantities. If the ratios of a / a1, b / b1, c / c1 are all about 3, and When the ratios of a / a1, b / b1, and c / c1 are all approximately 1, the straight line of the windsurfing board contour is correctly extracted and identified; if one of the ratios of a / a1, b / b1, and c / c1 is not approximately 3, or When one of the ratios is not approximately 1, it means that the sailboard contour straight line is incorrectly extracted and identified, and the results of the incorrect extraction and identification are eliminated; finally, the results of 8 groups of scaling ratios (feature side length / actual side length) and length parameter ratios (the ratio of two feature side lengths in a feature invariant / the ratio of two actual side lengths corresponding to the corresponding actual sailboard feature quantity) correctly extracted and identified are obtained as shown in Table 1 in the following table.
[0074] Table 1
[0075]
[0076] To sum up, the present invention provides a method for identifying characteristic parts of spatial non-cooperative targets, which forms a method for detecting the relative posture of spatial non-cooperative targets by capturing and identifying characteristic parts on the spatial non-cooperative targets; and has the characteristics of high precision and accurate measurement, and is particularly suitable for visual detection and identification of characteristic parts of spatial non-cooperative targets with high symmetry.
[0077] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the invention. After reading the above content, it will be obvious to those skilled in the art for various modifications and substitutions of the present invention. Therefore, the protection scope of the present invention should be limited by the attached claims.
Claims
1. A method for identifying characteristic parts of spatial non-cooperative targets, characterized in that: Include: S1, preprocessing the captured images of the spatial non-cooperative target feature parts and extracting lines in the captured images; S2, construct the closed spatial non-cooperative target feature part contour from the extracted image lines; S3, calculating multiple groups of feature invariants of multiple feature triangles selected from the contours of the feature parts of the spatial non-cooperative target; S4, identifying the correct feature part contour by judging all feature invariants of the spatial non-cooperative target feature parts.
2. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 1, characterized in that: The step S1 comprises: S1.1, performing contrast stretching preprocessing on the captured image of the spatial non-cooperative target feature part; S1.2, extracting lines from the captured image of the characteristic part of the spatial non-cooperative target using a straight line detection method, and obtaining parameters characterizing each line.
3. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 2, characterized in that: The straight line detection method is Hough transform or LSD straight line extraction method; the parameters characterizing each line are slope and intercept.
4. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 3, characterized in that: The step S2 comprises: S2.1, screening out characteristic straight lines through the positional relationship between the extracted lines, and determining contour straight lines of several spatial non-cooperative target characteristic parts from the characteristic straight lines; S2.2, connecting the contour straight lines end to end to obtain the contour of the closed spatial non-cooperative target feature part.
5. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 4, characterized in that: The screening of the characteristic straight lines includes: clustering the extracted lines and refitting them by judging the slope and intercept of each line; and judging the positional relationship between the intersections and endpoints of the fitted lines, and eliminating non-characteristic straight lines to obtain characteristic straight lines.
6. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 4, characterized in that: The step S3 comprises: S3.1, select any three adjacent contour lines from the contour of the spatial non-cooperative target feature part, and construct the first characteristic invariant of the first characteristic triangle based on the three contour lines; S3.2, construct the second characteristic triangle, the third characteristic triangle,..., the Nth characteristic triangle in turn for all the remaining contour lines, and for each characteristic triangle, construct the second characteristic invariant, the third characteristic invariant,..., the Nth characteristic invariant in turn.
7. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 6, characterized in that: The first characteristic variable is: [a / b, a / c, b / c, θ1, θ2, θ3]; Among them, a, b, c are the three characteristic side lengths of the first characteristic triangle, θ1, θ2, θ3 are the three included angles of the first characteristic triangle.
8. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 6, characterized in that: If any three adjacent contour lines cannot form a feature triangle, they will be skipped and not processed.
9. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 7, characterized in that: The step S4 comprises: S4.1, by using the difference in angle parameters in each set of feature invariants, all feature invariants of the obtained spatial non-cooperative target feature parts are compared one-to-one with the known feature quantities of the spatial non-cooperative target feature parts; S4.2, removing the feature invariants whose ratio of the scaling ratio to the length parameter exceeds a preset threshold, and obtaining the correct feature invariants, that is, obtaining the contour of the correctly extracted and identified feature part.
10. The method for identifying characteristic parts of spatial non-cooperative targets according to claim 9, characterized in that: The scaling ratio is: characteristic side length / actual side length; The ratio of the length parameters is: the ratio of two characteristic side lengths in a characteristic invariant / the ratio of two corresponding actual side lengths in the corresponding actual windsurfing board characteristic quantity.