A 3D target detection method based on laser point cloud

Abstract

The present invention provides a 3D target detection method based on laser point cloud, which relates to the technical field of target detection, and is used to solve the problem that errors are easy to occur in the 3D target detection process. Step S1: acquiring a picture of a target object, judging that the target object is a prismatic image or an arc image according to the picture, and generating measurement points at different positions according to images of different shapes; Step S2: establishing a three-dimensional rectangular coordinate system with the intersection of the leftmost measurement point of the target object and the parallel line at the bottom of the target object as the center o of the circle; Step S3: acquiring the coordinate points of the measurement points of the target object, connecting the coordinate points acquired through the front view, the rear view and the top view to form a three-dimensional object, and generating the three-dimensional object through an image generation unit; The present invention establishes a three-dimensional rectangular coordinate system based on the target object, represents the measurement points in the three-dimensional rectangular coordinate system, generates a 3D target object, and judges the target object according to the 3D image generated of the target object.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, and in particular to a 3D target detection method based on laser point cloud. Background Art

[0002] The laser radar includes a laser transmitter and a receiver. After the light emitted by the laser transmitter hits the target, it is reflected back to the receiver. The 3D spatial coordinates of each light point can be calculated based on the propagation speed of light. Each light point is a cloud point. In recent years, the 3D target detection technology of laser cloud points has been widely used in the fields of robots and unmanned driving, especially in the field of unmanned driving. In complex traffic environments, the detection of pedestrians or vehicles on the road is particularly important. Chinese patent page CN115171096 A discloses a 3D target detection method based on the fusion of RGB images and laser cloud points. In this patent, the RGB image to be detected is input into a 2D detection model for feature extraction, and then the image information is segmented based on the feature results to obtain a segmentation mask; the three-dimensional standard of the laser cloud point is projected to the two-dimensional image coordinates, and then a fixed number of virtual points are randomly generated according to the points where the cloud points are projected to the target segmentation mask, and then the virtual points are re-projected to the original point cloud, and then the fused point cloud enters the 3D target detection module. Compared with the 2D technology, although it improves the detection accuracy to a certain extent, its operation is complicated, and errors in any step will lead to inaccurate results.

[0003] Chinese patent CN 114627183 A discloses a laser point cloud 3D target detection method. A semantic segmentation branch is added to a 3D target detection algorithm model to obtain a 3D multi-task algorithm model. The semantic segmentation branch is trained using the semantic truth value generated after voxelization. The intermediate features of the semantic segmentation branch are refined and output to the detection branch to constrain the detection work of the detection branch. The scheme generates a pseudo image from the image by voxelization, and then uses the pseudo image for 2D convolution. Although the detection speed is fast, the accuracy is low. Therefore, the present invention proposes a 3D target detection method based on laser point cloud. Summary of the invention

[0004] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a 3D target detection method based on laser point cloud. The present invention is based on acquiring a picture of the target object, judging the shape of the target object based on the picture of the target object, generating measurement points in different forms according to different shapes, establishing a three-dimensional rectangular coordinate system, representing the measurement points in the three-dimensional rectangular coordinate system, selecting the front view, top view and rear view to outline the contour lines, generating a 3D target object, and judging the target object based on the generated 3D image of the target object.

[0005] In order to achieve the above object, the present invention is implemented by the following technical solution: a 3D target detection method based on laser point cloud, the detection method specifically comprises the following steps:

[0006] Step S1: Acquire a picture of the target object, determine whether the target object is a prismatic image or an arc-shaped image according to the picture of the front view, and generate measurement points at different positions according to images of different shapes;

[0007] Step S2: Taking the intersection of the leftmost measurement point of the target object and the parallel line at the bottom of the target object as the center o, a three-dimensional rectangular coordinate system is established, and the coordinate points are determined by using the length to represent the x-axis, the width to represent the y-axis, and the thickness to represent the z-axis;

[0008] Step S3: Acquire the coordinate points of the target object's measurement points, connect the coordinate points acquired through the front view, rear view and top view, connect them through straight lines if it is a prismatic image, and connect them through smooth curves if it is an arc image, to form a three-dimensional object, and generate the three-dimensional object through an image generation unit;

[0009] Step S4: The generated three-dimensional object is transmitted to the cloud server. The cloud server extracts the target object information stored in the storage module, compares it with the three-dimensional object, determines the target object information, and realizes the detection of the target object.

[0010] Furthermore, in step S1, if the intersection of the lines of the target object of the prism image is the measurement point, the number of measurement points is obtained; if it is an arc image, the position of the arc surface is determined, and a measurement point with respect to the arc image is obtained at intervals of 1 cm from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same.

[0011] Furthermore, in step S3, obtaining the coordinate points of the front view specifically includes the following steps:

[0012] Step S31: Count the number of intersections of the lines to obtain q points, where q=1, 2, 3, ..., k1, and k1 is a positive integer. Number each point. A laser emitter that can move along the x direction is provided on the top of the detection object. The laser emitter is moved to the top of the first number to form a vertical point with the x-axis. The distance between the vertical point and the center o of the circle is measured to obtain the measured distance xs1.

[0013] Step S32: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measured distance ys1;

[0014] Step S33: when the adjustment frame moves, there is a group of adjustment devices behind the detection object that keep moving synchronously with it, and the distance between the two groups of adjustment frames is obtained as L. A horizontal calibration surface is set in the middle of the two groups of adjustment frames. The horizontal calibration surface passes through the detection object and is located at the center of the thickness of the detection object, coinciding with the z-axis. The thickness measured before the detection object is in the negative direction of the z-axis, and the thickness measured after the detection object is in the negative direction of the z-axis.

[0015] Step S34: Measure the distance of the target object by a set of distance sensors to obtain L1, and obtain the measured distance zs1 by L / 2-L1, so that the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xs1, ys1, -zs1);

[0016] Step S35: gradually translate the laser emitters according to the sequence of numbers, gradually obtain the coordinate points of the second to z-th numbered points in the three-dimensional rectangular coordinate system, and obtain the coordinate point of the q-th point in the three-dimensional rectangular coordinate system as (xsq, ysq, -zsq);

[0017] Step S36: After the target object is determined to be acquired based on the image of the front view, the target object is determined based on the image of the rear view.

[0018] Furthermore, in step S36, obtaining the coordinate points of the rear view specifically includes the following steps:

[0019] Step S361: If the rear view is determined to be a prismatic image, the intersection of the lines of the target object is taken as the measurement point, and the number of measurement points is obtained; if it is an arc image, the arc surface position is determined, and a measurement point with an interval of 1 cm is obtained from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same;

[0020] Step S362: Count the number of line intersections to obtain q h points, q h =1, 2, 3...k2, k2 is a positive integer, each point is numbered, and a laser transmitter that can move along the x direction is set on the top of the detection object;

[0021] Step S363: Move the laser transmitter to the top of the first number to form a vertical point with the x-axis, measure the distance between the vertical point and the center o of the circle, and obtain the measured distance xl1;

[0022] Step S364: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measuring tool distance yl1;

[0023] Step S365: When another set of adjustment frames moves, another set of distance sensors is used to measure the distance of the target object to obtain L2, and the measured distance zl1 is obtained by L / 2-L2, so the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xl1, yl1, zl1);

[0024] Step S366: gradually translate the laser emitters according to the sequence of the numbers, gradually obtain the coordinate points of the second to z-th numbers in the three-dimensional rectangular coordinate system, and obtain the q-th h The coordinates of the point in the three-dimensional rectangular coordinate system are (xlq h , ylq h , zlq h ).

[0025] Furthermore, after the target object is determined by the image of the rear view, the target object is determined based on the image of the top view. If the top view is determined to be a prism-shaped convex image or an arc-shaped convex image, the coordinate points of the detection object are not acquired, and the coordinate points acquired by the front view and the rear view are connected. If it is a prism image, they are connected by straight lines. If it is an arc image, they are connected by smooth curves to form a three-dimensional object. The three-dimensional object is generated by the image generation unit. If it is determined to be a prism-shaped concave image, the intersection position of the lines of the target object is taken as the measurement point, and the number of measurement points is acquired.

[0026] Furthermore, obtaining the coordinate points of the top view specifically includes the following steps:

[0027] Step ①: If it is an arc-shaped concave image, a measurement point is obtained from the top to the bottom and from the left to the right at an interval of 1 cm with respect to the arc-shaped image, and the measurement distances of multiple arc-shaped images obtained from the top to the bottom are the same;

[0028] Step ②: Measure the distance between the two measuring points on the top side through the image to obtain the image distance ts, obtain the coordinate points through the front view and the rear view, and display the obtained coordinate points in the image to determine the position of the corresponding coordinate points;

[0029] Step 3: Obtain the distance between the two coordinate points according to the positions of the coordinate points to obtain the actual distance sjs, and obtain the distance ratio jb through the ratio between the actual distance and the image distance;

[0030] Step ④: Obtain the coordinate points of the detected object on the image, connect the two sets of coordinate points to form a straight line, define the obtained straight line as a reference straight line, connect an intersection point of the reference straight line with a measurement point at the edge of the concave position, and obtain the measurement line distance;

[0031] Step ⑤: Measure the distance of the measuring line and compare the obtained distance with the actual distance sj of the measuring line l Acquiring, in the process of generating the measuring line, acquiring the angle between the measuring line and the reference straight line;

[0032] Step ⑥: Draw a line segment perpendicular to the reference line at the measuring point to obtain the vertical point. Use trigonometric functions to calculate the distance between the reference line coordinate point and the vertical point. The distance Jx is the horizontal distance based on the reference point on the x-axis.

[0033] Step 7: Set the reference point to (cx, cy, cz), that is, the position of the vertical point is (cx±Jx, cy, cz), obtain the vertical distance between the reference line segment and the measuring line based on the position of the vertical point, and use trigonometric functions to calculate the vertical distance Jy. The obtained distance Jy is the horizontal distance on the y-axis based on the reference point;

[0034] Step ⑧: A movable distance measuring instrument is provided on the detection object, and the distance to the coordinate point is obtained by the distance measuring instrument. The height of the distance measuring instrument is determined according to the position of the coordinate point, and the distance measuring instrument is moved to the top of the measurement point to obtain the distance to the measurement point;

[0035] Step ⑨: Determine the moving distance of the measuring point on the z-axis as Jz based on the difference between the measuring point and the coordinate point, thereby obtaining the coordinate point of the measuring point in the three-dimensional rectangular coordinate system as (cx±Jx, cy±Jy, cz±Jz);

[0036] Step ⑩: Connect the positions of multiple measurement points on the top view, if it is a prismatic image, connect them through straight lines, if it is an arc image, connect them through smooth curves, to form a three-dimensional object, and generate the three-dimensional object through the image generation unit.

[0037] Furthermore, when storing the target objects, the storage module classifies the target objects according to their shape information, and classifies them according to irregular shapes, regular shapes, and circles. When searching for the target objects, a quick search is performed based on the classification.

[0038] Beneficial effects of the present invention:

[0039] 1. The present invention is based on acquiring a picture of a target object, judging the shape of the target object based on the picture of the target object, generating measurement points in different forms according to different shapes, establishing a three-dimensional rectangular coordinate system, representing the measurement points in the three-dimensional rectangular coordinate system, selecting a front view, a top view and a rear view to outline the contour lines, generating a 3D target object, and judging the target object based on the generated 3D image of the target object.

[0040] 2. The present invention analyzes the top image of the target object and makes a judgment based on the top image during the generation process of the target object. The position of the measuring point in the three-dimensional rectangular coordinate system is determined by performing a comprehensive analysis based on the acquired coordinate points and the unacquired measuring points, thereby improving the convenience of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0042] Figure 1 This is a principle block diagram of a 3D target detection method based on laser point cloud of the present invention;

[0043] Figure 2 A step diagram of a 3D target detection method based on laser point cloud according to the present invention;

[0044] Figure 3 This is a step diagram of a 3D target detection method based on laser point cloud of the present invention. DETAILED DESCRIPTION

[0045] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0046] In the present invention, please refer to Figure 1-Figure 3 , a 3D target detection method based on laser point cloud, the detection method includes a stage adjustment device, a picture acquisition unit, a laser calibration unit, an image generation unit, a storage module and a cloud server;

[0047] The industry name for cloud servers is actually computing units. The so-called computing unit means that this server can only be regarded as a person's brain, equivalent to the CPU of an ordinary computer, and the resources inside are limited. To obtain better performance, the solution is to upgrade the cloud server or deploy other software that consumes computing unit resources on the corresponding cloud service. For example, there are dedicated cloud database services for databases, and dedicated file storage services for static web pages and images.

[0048] The adjustment equipment includes a measuring instrument, an adjustment frame and a distance sensor;

[0049] The details are as follows:

[0050] The image of the target object is acquired by the image acquisition unit, and the target object is judged to be a prismatic image or an arc image according to the image of the front view. If it is a prismatic image, the intersection of the lines of the target object is the measurement point, and the number of measurement points is acquired;

[0051] If it is an arc image, to determine the position of the arc surface, a measurement point with respect to the arc image is obtained at intervals of 1 cm from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same;

[0052] Take the intersection of the leftmost measurement point of the target object and the parallel line at the bottom of the target object as the center o, establish a three-dimensional rectangular coordinate system, and use length to represent the x-axis, width to represent the y-axis, and thickness to represent the z-axis to determine the coordinate points;

[0053] The number of intersections of the lines is counted to obtain q points, q=1, 2, 3...k1, k1 is a positive integer, and each point is numbered. A laser emitter that can move along the x direction is set on the top of the test object. The laser emitter moves to the top of the first number to form a vertical point with the x-axis. The distance between the vertical point and the center o of the circle is measured to obtain the measured distance xs1;

[0054] After measuring the length of the first measuring point, the laser in the laser calibration unit generates laser light, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measured distance ys1;

[0055] When the adjustment frame moves, there is a set of adjustment devices that keep moving synchronously with the detection object behind it. The distance between the two sets of adjustment frames is obtained as L. A horizontal calibration plane is set in the middle of the two sets of adjustment frames. The horizontal calibration plane passes through the detection object and is located at the center of the thickness of the detection object. It coincides with the z-axis. The thickness measured before the detection object is in the negative direction of the z-axis, and the thickness measured after the detection object is in the negative direction of the z-axis. The distance of the target object is measured by a set of distance sensors to obtain L1. The measured distance zs1 is obtained by L / 2-L1, so the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xs1, ys1, -zs1);

[0056] The laser emitters are gradually translated according to the sequence of numbers, and the coordinate points of the second to z-th numbers in the three-dimensional rectangular coordinate system are gradually obtained, and the coordinate point of the q-th point in the three-dimensional rectangular coordinate system is (xsq, ysq, -zsq);

[0057] After the target object is determined to be acquired according to the front view image, the target object is determined according to the rear view image. If the rear view image is a prismatic image, the intersection of the lines of the target object is taken as the measurement point, and the number of measurement points is obtained;

[0058] If it is an arc image, to determine the position of the arc surface, a measurement point with respect to the arc image is obtained at intervals of 1 cm from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same;

[0059] Count the number of line intersections and get q h points, q h =1, 2, 3...k2, k2 is a positive integer, each point is numbered, a laser emitter that can move along the x direction is set on the top of the test object, the laser emitter is moved to the top of the first number, and a vertical point is formed with the x-axis, and the distance between the vertical point and the center o is measured to obtain the measured distance xl1;

[0060] After measuring the length of the first measuring point, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measuring tool distance yl1;

[0061] When another set of adjustment frames moves, another set of distance sensors is used to measure the distance of the target object to obtain L2, and the measured distance zl1 is obtained by L / 2-L2, so the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xl1, yl1, zl1);

[0062] The laser emitters are gradually translated according to the sequence of numbers, and the coordinate points of the second to z-th numbers in the three-dimensional rectangular coordinate system are gradually acquired to obtain the q-th h The coordinates of the point in the three-dimensional rectangular coordinate system are (xlq h , ylq h , zlq h );

[0063] After the target object is determined by the rear view image, the target object is determined according to the top view image. If the top view image is a prism-shaped convex image or an arc-shaped convex image, the coordinate points of the detected object are not obtained, and the coordinate points obtained by the front view and the rear view are connected. If it is a prism-shaped image, they are connected by straight lines. If it is an arc-shaped image, they are connected by smooth curves to form a three-dimensional object. The three-dimensional object is generated by the image generation unit;

[0064] If it is judged to be a prismatic concave image, the intersection of the lines of the target object is taken as the measurement point, and the number of measurement points is obtained;

[0065] If it is an arc-shaped concave image, a measurement point on the arc-shaped image is obtained at intervals of 1 cm from top to bottom and from left to right, and the measurement distances of multiple arc-shaped images obtained from top to bottom are the same;

[0066] The distance between the two measuring points on the top side is measured through the picture to obtain the picture distance ts. The coordinate points are obtained through the front view and the rear view, and the obtained coordinate points are displayed in the picture. The positions of the corresponding coordinate points are judged, and the distance between the two coordinate points is obtained according to the positions of the coordinate points to obtain the actual distance sjs. The ratio between the actual distance and the picture distance is obtained to obtain the distance ratio jb. The coordinate points of the detected object obtained on the picture are obtained, and the two groups of coordinate points are connected to form a straight line. The obtained straight line is defined as the reference straight line. An intersection point of the reference straight line is connected with a measuring point at the edge of the concave position to obtain the measuring line distance. The measuring line distance is measured, and the actual distance sj of the measuring line is compared with the obtained distance. l To obtain, in the process of generating the measuring line, obtain the angle between the measuring line and the reference line, draw a line segment perpendicular to the reference line at the measuring point, obtain the vertical point, and calculate the distance between the reference line coordinate point and the vertical point through trigonometric functions, and the distance Jx is the horizontal distance based on the reference point on the x-axis;

[0067] The trigonometric function is used to calculate the angle as follows: Set the angle to α, cosα=Jx / sj l ;

[0068] Set the reference point to (cx, cy, cz), that is, the position of the vertical point is (cx±Jx, cy, cz), obtain the vertical distance between the reference line segment and the measuring line based on the position of the vertical point, and calculate the vertical distance Jy through trigonometric function. The obtained distance Jy is the horizontal distance based on the reference point on the y-axis;

[0069] A movable distance measuring instrument is provided on the detection object, and the distance between the distance measuring instrument and the coordinate point is obtained by the distance measuring instrument. The height of the distance measuring instrument is determined according to the position of the coordinate point, and the distance between the distance measuring instrument and the measurement point is obtained. The moving distance of the measurement point on the z-axis at the coordinate point is determined to be Jz according to the difference between the distance measuring instrument and the coordinate point, thereby obtaining the coordinate point of the measurement point in the three-dimensional rectangular coordinate system as (cx±Jx, cy±Jy, cz±Jz);

[0070] Thus, the positions of multiple measurement points on the top view are connected, if it is a prismatic image, they are connected by straight lines, if it is an arc image, they are connected by smooth curves, to form a three-dimensional object, and the three-dimensional object is generated by the image generation unit;

[0071] The generated three-dimensional object is transmitted to the cloud server, and the cloud service extracts the target object information stored in the storage module, compares it with the three-dimensional object, determines the target object information, and realizes the detection of the target object.

[0072] It should be noted that when the storage module stores the target objects, it classifies them according to their shape information, i.e., they are classified into irregular shapes, regular shapes, and circles. When searching for the target objects, they are quickly searched based on the classification.

[0073] In another embodiment, based on another concept of the same invention, a working method of a 3D target detection method based on laser point cloud is proposed, and the specific steps of the working method are as follows:

[0074] Step S1: Obtain a picture of the target object, and determine whether the target object is a prismatic image or an arc image according to the picture of the front view. If it is a prismatic image, the intersection of the lines of the target object is the measurement point, and the number of measurement points is obtained; if it is an arc image, determine the position of the arc surface, and obtain a measurement point with an interval of 1 cm from top to bottom and from left to right. The measurement distances of multiple arc images obtained from top to bottom are the same;

[0075] Step S2: Taking the intersection of the leftmost measurement point of the target object and the parallel line at the bottom of the target object as the center o, a three-dimensional rectangular coordinate system is established, and the coordinate points are determined by using the length to represent the x-axis, the width to represent the y-axis, and the thickness to represent the z-axis;

[0076] Step S3: Acquire the coordinate points of the target object's measurement points, connect the coordinate points acquired through the front view, rear view and top view, connect them through straight lines if it is a prismatic image, and connect them through smooth curves if it is an arc image, to form a three-dimensional object, and generate the three-dimensional object through an image generation unit;

[0077] The acquisition of the coordinate points of the front view includes the following steps:

[0078] Step S31: Count the number of intersections of the lines to obtain q points, where q=1, 2, 3, ..., k1, and k1 is a positive integer. Number each point. A laser emitter that can move along the x direction is provided on the top of the detection object. The laser emitter is moved to the top of the first number to form a vertical point with the x-axis. The distance between the vertical point and the center o of the circle is measured to obtain the measured distance xs1.

[0079] Step S32: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measured distance ys1;

[0080] Step S33: when the adjustment frame moves, there is a group of adjustment devices behind the detection object that keep moving synchronously with it, and the distance between the two groups of adjustment frames is obtained as L. A horizontal calibration surface is set in the middle of the two groups of adjustment frames. The horizontal calibration surface passes through the detection object and is located at the center of the thickness of the detection object, coinciding with the z-axis. The thickness measured before the detection object is in the negative direction of the z-axis, and the thickness measured after the detection object is in the negative direction of the z-axis.

[0081] Step S34: Measure the distance of the target object by a set of distance sensors to obtain L1, and obtain the measured distance zs1 by L / 2-L1, so that the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xs1, ys1, -zs1);

[0082] Step S35: gradually translate the laser emitters according to the sequence of numbers, gradually obtain the coordinate points of the second to z-th numbered points in the three-dimensional rectangular coordinate system, and obtain the coordinate point of the q-th point in the three-dimensional rectangular coordinate system as (xsq, ysq, -zsq);

[0083] Step S36: after the target object is determined to be acquired according to the front view image, the target object is determined according to the rear view image;

[0084] The acquisition of the coordinate points of the rear view includes the following steps:

[0085] Step S361: If the rear view is determined to be a prismatic image, the intersection of the lines of the target object is taken as the measurement point, and the number of measurement points is obtained; if it is an arc image, the arc surface position is determined, and a measurement point with an interval of 1 cm is obtained from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same;

[0086] Step S362: Count the number of line intersections to obtain q h points, q h =1, 2, 3...k2, k2 is a positive integer, each point is numbered, and a laser transmitter that can move along the x direction is set on the top of the detection object;

[0087] Step S363: Move the laser transmitter to the top of the first number to form a vertical point with the x-axis, measure the distance between the vertical point and the center o of the circle, and obtain the measured distance xl1;

[0088] Step S364: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measuring tool distance yl1;

[0089] Step S365: When another set of adjustment frames moves, another set of distance sensors is used to measure the distance of the target object to obtain L2, and the measured distance zl1 is obtained by L / 2-L2, so the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xl1, yl1, zl1);

[0090] Step S366: gradually translate the laser emitters according to the sequence of the numbers, gradually obtain the coordinate points of the second to z-th numbers in the three-dimensional rectangular coordinate system, and obtain the q-th h The coordinates of the point in the three-dimensional rectangular coordinate system are (xlq h , ylq h , zlq h );

[0091] Step S367: After the target object is determined by the rear view image, the target object is determined according to the top view image. If the top view image is a prism-shaped convex image or an arc-shaped convex image, the coordinate points of the detection object are not acquired, and the coordinate points acquired by the front view and the rear view are connected. If it is a prism-shaped image, they are connected by straight lines, and if it is an arc-shaped image, they are connected by smooth curves to form a three-dimensional object, and the three-dimensional object is generated by the image generation unit;

[0092] Step S368: If it is determined to be a prismatic concave image, the intersection of the lines of the target object is taken as the measurement point, and the number of the measurement points is obtained;

[0093] The acquisition of the coordinate points of the top view includes the following steps:

[0094] Step ①: If it is an arc-shaped concave image, a measurement point is obtained from the top to the bottom and from the left to the right at an interval of 1 cm with respect to the arc-shaped image, and the measurement distances of multiple arc-shaped images obtained from the top to the bottom are the same;

[0095] Step ②: Measure the distance between the two measuring points on the top side through the image to obtain the image distance ts, obtain the coordinate points through the front view and the rear view, and display the obtained coordinate points in the image to determine the position of the corresponding coordinate points;

[0096] Step 3: Obtain the distance between the two coordinate points according to the positions of the coordinate points to obtain the actual distance sjs, and obtain the distance ratio jb through the ratio between the actual distance and the image distance;

[0097] Step ④: Obtain the coordinate points of the detected object on the image, connect the two sets of coordinate points to form a straight line, define the obtained straight line as a reference straight line, connect an intersection point of the reference straight line with a measurement point at the edge of the concave position, and obtain the measurement line distance;

[0098] Step ⑤: Measure the distance of the measuring line and compare the obtained distance with the actual distance sj of the measuring line l Acquiring, in the process of generating the measuring line, acquiring the angle between the measuring line and the reference straight line;

[0099] Step ⑥: Draw a line segment perpendicular to the reference line at the measuring point to obtain the vertical point. Use trigonometric functions to calculate the distance between the reference line coordinate point and the vertical point. The distance Jx is the horizontal distance based on the reference point on the x-axis.

[0100] Step 7: Set the reference point to (cx, cy, cz), that is, the position of the vertical point is (cx±Jx, cy, cz), obtain the vertical distance between the reference line segment and the measuring line based on the position of the vertical point, and use trigonometric functions to calculate the vertical distance Jy. The obtained distance Jy is the horizontal distance on the y-axis based on the reference point;

[0101] Step ⑧: A movable distance measuring instrument is provided on the detection object, and the distance to the coordinate point is obtained by the distance measuring instrument. The height of the distance measuring instrument is determined according to the position of the coordinate point, and the distance measuring instrument is moved to the top of the measurement point to obtain the distance to the measurement point;

[0102] Step ⑨: Determine the moving distance of the measuring point on the z-axis as Jz based on the difference between the measuring point and the coordinate point, thereby obtaining the coordinate point of the measuring point in the three-dimensional rectangular coordinate system as (cx±Jx, cy±Jy, cz±Jz);

[0103] Step ⑩: thereby connecting the positions of multiple measurement points on the top view, if it is a prismatic image, they are connected by straight lines, if it is an arc image, they are connected by smooth curves, to form a three-dimensional object, and the three-dimensional object is generated by an image generation unit;

[0104] Step S4: The generated three-dimensional object is transmitted to the cloud server. The cloud server extracts the target object information stored in the storage module, compares it with the three-dimensional object, determines the target object information, and realizes the detection of the target object.

[0105] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technical personnel in this field according to actual conditions. For example, if there are weight coefficients and proportional coefficients, their set sizes are to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. Regarding the size of the weight coefficient and the proportional coefficient, it is sufficient as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0106] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0107] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0108] The above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A 3D target detection method based on laser point cloud, characterized in that: The detection method specifically comprises the following steps: Step S1: Acquire a picture of the target object, determine whether the target object is a prismatic image or an arc-shaped image according to the picture of the front view, and generate measurement points at different positions according to images of different shapes; Step S2: Taking the intersection of the leftmost measurement point of the target object and the parallel line at the bottom of the target object as the center o, a three-dimensional rectangular coordinate system is established, and the coordinate points are determined by using the length to represent the x-axis, the width to represent the y-axis, and the thickness to represent the z-axis; Step S3: Acquire the coordinate points of the target object's measurement points, connect the coordinate points acquired through the front view, rear view and top view, connect them through straight lines if it is a prismatic image, and connect them through smooth curves if it is an arc image, to form a three-dimensional object, and generate the three-dimensional object through an image generation unit; In the step S3, obtaining the coordinate points of the front view specifically includes the following steps: Step S31: Count the number of intersections of the lines to obtain q points, where q=1, 2, 3, ..., k1, and k1 is a positive integer. Number each point. A laser emitter that can move along the x direction is provided on the top of the detection object. The laser emitter is moved to the top of the first number to form a vertical point with the x-axis. The distance between the vertical point and the center o of the circle is measured to obtain the measured distance xs1. Step S32: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measured distance ys1; Step S33: when the adjustment frame moves, there is a group of adjustment devices behind the detection object that keep moving synchronously with it, and the distance between the two groups of adjustment frames is obtained as L. A horizontal calibration surface is set in the middle of the two groups of adjustment frames. The horizontal calibration surface passes through the detection object and is located at the center of the thickness of the detection object, coinciding with the z-axis. The thickness measured before the detection object is in the negative direction of the z-axis, and the thickness measured after the detection object is in the negative direction of the z-axis. Step S34: Measure the distance of the target object by a set of distance sensors to obtain L1, and obtain the measured distance zs1 by L / 2-L1, so that the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xs1, ys1, -zs1); Step S35: gradually translate the laser emitters according to the sequence of numbers, gradually obtain the coordinate points of the second to z-th numbered points in the three-dimensional rectangular coordinate system, and obtain the coordinate point of the q-th point in the three-dimensional rectangular coordinate system as (xsq, ysq, -zsq); Step S36: after the target object is determined to be acquired according to the front view image, the target object is determined according to the rear view image; Step S4: The generated three-dimensional object is transmitted to the cloud server. The cloud server extracts the target object information stored in the storage module, compares it with the three-dimensional object, determines the target object information, and realizes the detection of the target object.

2. The 3D target detection method based on laser point cloud according to claim 1, characterized in that: In step S1, if the intersection of the lines of the target object of the prism image is the measurement point, the number of measurement points is obtained; if it is an arc image, the position of the arc surface is determined, and a measurement point with respect to the arc image is obtained at intervals of 1 cm from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same.

3. The 3D target detection method based on laser point cloud according to claim 1, characterized in that: In the step S36, obtaining the coordinate points of the rear view specifically includes the following steps: Step S361: If the rear view is determined to be a prismatic image, the intersection of the lines of the target object is taken as the measurement point, and the number of measurement points is obtained; if it is an arc image, the arc surface position is determined, and a measurement point with an interval of 1 cm is obtained from top to bottom and from left to right, and the measurement distances of multiple arc images obtained from top to bottom are the same; Step S362: Count the number of intersections of the lines to obtain qh points, where qh=1, 2, 3, ..., k2, and k2 is a positive integer. Number each point. A laser transmitter that can move along the x direction is provided on the top of the detection object. Step S363: Move the laser transmitter to the top of the first number to form a vertical point with the x-axis, measure the distance between the vertical point and the center o of the circle, and obtain the measured distance xl1; Step S364: After the length of the first measuring point is measured, the laser in the laser calibration unit generates laser, and the laser is moved to the vertical point through the adjustment frame, and the laser is moved upward so that the laser emitted by the laser coincides with the measuring point. The measuring instrument measures the moving distance to obtain the measuring tool distance yl1; Step S365: When another set of adjustment frames moves, another set of distance sensors is used to measure the distance of the target object to obtain L2, and the measured distance zl1 is obtained by L / 2-L2, so the coordinate point of the first point in the three-dimensional rectangular coordinate system is (xl1, yl1, zl1); Step S366: gradually translate the laser emitter in order of the numbers, gradually obtain the coordinate points of the second to zth numbered points in the three-dimensional rectangular coordinate system, and obtain the coordinate point of the qhth point in the three-dimensional rectangular coordinate system as (xlqh, ylqh, zlqh).

4. The 3D target detection method based on laser point cloud according to claim 3, characterized in that: After the target object is acquired by judging the image of the rear view, the target object is judged according to the image of the top view. If the top view is judged to be a prism-shaped convex image or an arc-shaped convex image, the coordinate points of the detection object are not acquired. The coordinate points acquired by the front view and the rear view are connected. If it is a prism-shaped image, they are connected by straight lines. If it is an arc-shaped image, they are connected by smooth curves to form a three-dimensional object. The three-dimensional object is generated by the image generation unit. If it is judged to be a prism-shaped concave image, the intersection position of the lines of the target object is the measurement point, and the number of measurement points is acquired.

5. The 3D target detection method based on laser point cloud according to claim 4, characterized in that: The acquisition of the coordinate points of the top view includes the following steps: Step ①: If it is an arc-shaped concave image, a measurement point is obtained from the top to the bottom and from the left to the right at an interval of 1 cm with respect to the arc-shaped image, and the measurement distances of multiple arc-shaped images obtained from the top to the bottom are the same; Step ②: Measure the distance between the two measuring points on the top side through the image to obtain the image distance ts, obtain the coordinate points through the front view and the rear view, and display the obtained coordinate points in the image to determine the position of the corresponding coordinate points; Step 3: Obtain the distance between the two coordinate points according to the positions of the coordinate points to obtain the actual distance sjs, and obtain the distance ratio jb through the ratio between the actual distance and the image distance; Step ④: Obtain the coordinate points of the detected object on the image, connect the two sets of coordinate points to form a straight line, define the obtained straight line as a reference straight line, connect an intersection point of the reference straight line with a measurement point at the edge of the concave position, and obtain the measurement line distance; Step ⑤: Measure the distance of the measuring line, obtain the actual distance sjl of the measuring line by comparing the obtained distance, and obtain the angle between the measuring line and the reference straight line during the process of generating the measuring line; Step ⑥: Draw a line segment perpendicular to the reference line at the measuring point to obtain the vertical point. Use trigonometric functions to calculate the distance between the reference line coordinate point and the vertical point. The distance Jx is the horizontal distance based on the reference point on the x-axis. Step 7: Set the reference point to (cx, cy, cz), that is, the position of the vertical point is (cx±Jx, cy, cz), obtain the vertical distance between the reference line segment and the measuring line based on the position of the vertical point, and use trigonometric functions to calculate the vertical distance Jy. The obtained distance Jy is the horizontal distance on the y-axis based on the reference point; Step ⑧: A movable distance measuring instrument is provided on the detection object, and the distance to the coordinate point is obtained by the distance measuring instrument. The height of the distance measuring instrument is determined according to the position of the coordinate point, and the distance measuring instrument is moved to the top of the measurement point to obtain the distance to the measurement point; Step ⑨: Determine the moving distance of the measuring point on the z-axis as Jz based on the difference between the measuring point and the coordinate point, thereby obtaining the coordinate point of the measuring point in the three-dimensional rectangular coordinate system as (cx±Jx, cy±Jy, cz±Jz); Step ⑩: Connect the positions of multiple measurement points on the top view, if it is a prismatic image, connect them through straight lines, if it is an arc image, connect them through smooth curves, to form a three-dimensional object, and generate the three-dimensional object through the image generation unit.

6. The 3D target detection method based on laser point cloud according to claim 1, characterized in that: When storing the target objects, the storage module classifies the target objects according to their shape information, i.e., classifies them according to irregular shapes, regular shapes, and circles. When searching for the target objects, a quick search is performed based on the classification.

Images