Dynamic object moving distance measuring method and system based on laser point cloud, terminal and medium

Abstract

The invention discloses a dynamic object moving distance measurement method and system based on laser point cloud, a terminal and a medium, and relates to the technical field of point cloud analysis, and the method comprises the steps: obtaining the current frame point cloud data of a moving target, and carrying out the registration of the current frame point cloud data with a target point cloud template sequence, and obtaining an optimal matching template; moving the optimal matching template to a target position corresponding to the point cloud data of the current frame to obtain a reference point of the point cloud data of the current frame; and judging whether the current frame point cloud data is the end frame point cloud data, and if the current frame point cloud data is the end frame point cloud data, determining the moving distance of the moving target. According to the method, the optimal matching template is searched for to realize coarse registration, and then the optimal matching template is moved to the target position corresponding to the point cloud data of the current frame to realize fine registration, so that accurate positioning of a moving target is realized, accurate measurement of the moving distance of a dynamic object can be realized, and the method has high robustness.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of point cloud analysis, and particularly relates to a dynamic object moving distance measurement method and system based on laser point cloud, a terminal and a medium. BACKGROUND

[0002] Measuring the actual pushing distance of a hydraulic support in a coal mine has been a technical pain point in the coal industry. The manual measurement method is dangerous, and personnel need to be exposed to the coal mining environment in the mine, and the movement of the hydraulic machine can easily cause harm to personnel. The moving distance of the hydraulic support is related to the production safety of coal mining, and an automatic measurement method is urgently needed.

[0003] However, the use of a laser radar for measurement requires processing and analyzing laser radar data, comparing the distance of the marker in the laser radar data before and after movement, and thus measuring the length. However, the marker of the laser radar is very unstable in the mine, which is due to the fact that the marker can be covered by an object or can be blocked by a dust layer, making it difficult to be completely or accurately collected by the laser radar. Therefore, in a high-noise environment, the use of traditional SLAM (Simultaneous localization and mapping), point cloud registration, and target recognition techniques for object movement analysis based on laser point cloud is affected by the high noise of the laser radar, and these algorithms cannot accurately identify and distinguish the marker through single-frame point cloud data, resulting in inaccurate measurement of the moving distance.

[0004] Therefore, the prior art still has defects. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a dynamic object moving distance measurement method and system based on laser point cloud, a terminal and a medium to solve the above-mentioned defects of the prior art. The technical solution adopted by the present application is as follows: In a first aspect, the present application provides a dynamic object moving distance measurement method based on laser point cloud, wherein the method comprises: acquiring current frame point cloud data of a moving target, registering the current frame point cloud data with a preset target point cloud template sequence to obtain a best matching template; moving the best matching template to a target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data; based on the reference point of the current frame point cloud data, determining whether the current frame point cloud data is end frame point cloud data, and if the current frame point cloud data is end frame point cloud data, based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data, obtaining the moving distance of the moving target.

[0006] In an implementation manner, the motion target is a landmark component or a landmark position on the scanned device, and when the method is applied to the measurement of the moving distance of the downhole hydraulic support, the motion target is the front head position of the working platform of the downhole hydraulic support.

[0007] In an implementation manner, the target point cloud template sequence is generated in the following manner: scanning the movement of the motion target in the whole motion cycle to obtain a point cloud flow of the motion target; converting the point cloud flow of the motion target into single-frame point cloud data for storage, and segmenting the single-frame point cloud data to obtain point cloud data of the motion target in each frame; obtaining the target point cloud template sequence according to the distribution of the point cloud data of the motion target in different frames in time and space.

[0008] In an implementation manner, the current frame point cloud data is registered with the preset target point cloud template sequence to obtain the best matching template, including: obtaining a point cloud data set of each template in the target point cloud template sequence; calculating a one-way Hausdorff distance between the current frame point cloud data and the point cloud data set of each template; taking the template corresponding to the minimum one-way Hausdorff distance as the best matching template.

[0009] In an implementation manner, the best matching template is moved to a target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data, including: determining the highest point of the best matching template, determining the closest point in the current frame point cloud data to the highest point, and moving the best matching template and the corresponding highest point to the closest point to obtain a point cloud data set of the best matching template after the first movement; taking the highest point of the best matching template after the first movement as a starting point, determining the highest points within a certain range in the current frame point cloud data, and continuing to move the best matching template after the first movement to the highest points within the certain range to obtain a point cloud data set of the best matching template after the second movement; respectively calculating the Hausdorff distances between the current frame point cloud data and the point cloud data set of the best matching template, the point cloud data set of the best matching template after the first movement, and the point cloud data set of the best matching template after the second movement; obtaining the reference point of the current frame point cloud data based on the minimum Hausdorff distance.

[0010] In an implementation manner, the determining whether the current frame point cloud data is end frame point cloud data based on the reference point of the current frame point cloud data comprises: calculating an Euclidean distance between the reference point of the current frame point cloud data and a reference point of initial frame point cloud data; determining whether the Euclidean distance meets a motion end condition based on the Euclidean distance and a distance mean calculated according to a time window in advance; if the motion end condition is met, determining that the current frame point cloud data is end frame point cloud data.

[0011] In an implementation manner, the method further comprises: performing point-by-point analysis on all points in the current frame point cloud data to obtain a best reference point of the current frame point cloud data, and obtaining a moving distance of the moving target based on the best reference point.

[0012] In a second aspect, the embodiments of the present application further provide a dynamic object moving distance measurement system based on laser point cloud, wherein the system comprises: a coarse registration module configured to acquire current frame point cloud data of a moving target, register the current frame point cloud data with a preset target point cloud template sequence, and obtain a best matching template; a fine registration module configured to move the best matching template to a target position corresponding to the current frame point cloud data, and obtain a reference point of the current frame point cloud data; a distance determination module configured to determine whether the current frame point cloud data is end frame point cloud data based on the reference point of the current frame point cloud data, and if the current frame point cloud data is end frame point cloud data, obtain a moving distance of the moving target based on a reference point of initial frame point cloud data and the reference point of the current frame point cloud data.

[0013] In a third aspect, the embodiments of the present application further provide a terminal, wherein the terminal comprises a memory, a processor, and a dynamic object moving distance measurement program based on laser point cloud stored in the memory and executable on the processor, and the processor implements the steps of the dynamic object moving distance measurement method based on laser point cloud in any of the above-mentioned solutions when executing the dynamic object moving distance measurement program based on laser point cloud.

[0014] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, wherein the computer readable storage medium stores a dynamic object moving distance measurement program based on laser point cloud, and the dynamic object moving distance measurement program based on laser point cloud implements the steps of the dynamic object moving distance measurement method based on laser point cloud in any of the above-mentioned solutions on the computer readable storage medium.

[0015] Beneficial effect: Compared with the prior art, the present invention provides a method for measuring the moving distance of a dynamic object based on laser point cloud, which first obtains the current frame point cloud data of the moving target, aligns the current frame point cloud data with a preset target point cloud template sequence, and obtains the best matching template. Then, the best matching template is moved to the target position corresponding to the current frame point cloud data to obtain the reference point of the current frame point cloud data. Next, based on the reference point of the current frame point cloud data, it is determined whether the current frame point cloud data is the end frame point cloud data. If the current frame point cloud data is the end frame point cloud data, the moving distance of the moving target is obtained based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data. The present invention realizes coarse alignment by finding the best matching template, and then moves the best matching template to the target position corresponding to the current frame point cloud data to realize fine alignment, thereby realizing precise positioning of the moving target, and can realize accurate measurement of the moving distance of the dynamic object, with high robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention provides a flowchart of a preferred embodiment of a method for measuring the moving distance of a dynamic object based on laser point cloud.

[0017] Figure 2 This is a technical roadmap for a method for measuring the moving distance of dynamic objects based on laser point clouds provided in an embodiment of the present invention.

[0018] Figure 3 A schematic diagram of a target point cloud template sequence and a single template in a method for measuring the moving distance of a dynamic object based on laser point cloud provided in an embodiment of the present invention.

[0019] Figure 4 A diagram illustrating the process of moving the best matching template in the method for measuring the moving distance of a dynamic object based on laser point cloud provided in an embodiment of the present invention.

[0020] Figure 5 A schematic diagram of a precise registration example in a method for measuring the moving distance of a dynamic object based on laser point cloud provided in an embodiment of the present invention.

[0021] Figure 6 This is a principle block diagram of a dynamic object moving distance measurement system based on laser point cloud provided by an embodiment of the present invention.

[0022] Figure 7 This is a functional block diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] For the purposes of the present application, the technical solutions and effects, the following will be further described in detail with reference to the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not intended to limit the present application.

[0024] The flowchart shown in the drawings is only an example and does not necessarily include all contents and operations or steps, nor does it necessarily execute in the order described. For example, some operations or steps can be further divided, combined or partially merged, so the actual execution order can be changed according to the actual situation.

[0025] It should be understood that the terms used in the present application specification herein are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, unless otherwise clear from the context, the singular forms "a", "an" and "the" are intended to include the plural forms. It should be understood that in order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, the same items or similar items with basically the same function and role are distinguished by using "first", "second" and the like. For example, the first control information and the second control information are only used to distinguish different control information, and do not limit the order. Those skilled in the art can understand that the "first", "second" and the like do not limit the quantity and execution order, and the "first", "second" and the like do not necessarily mean different. It should also be understood that the term "and / or" used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0026] The embodiment is based on laser radar to measure the moving distance of dynamic objects, such as the advancing degree of hydraulic support. However, the current measurement of the advancing degree of the underground hydraulic support mainly adopts the traditional manual measurement method, including visual measurement and ruler measurement, and lacks an automatic measurement method. The laser radar actively collects the environment and can work in a dark environment, and the measured data is a three-dimensional point set data, which accurately records the three-dimensional coordinates of the scanned object, and describes the surrounding environment with points instead of surfaces. However, the current use of laser radar for real-time advancing degree measurement of hydraulic support faces great technical challenges. The reasons are as follows: (1) Large dust noise in underground. The working environment of coal mining underground has a large amount of dust, and the laser radar itself is sensitive to the dust layer, which is easy to produce a large number of discrete flying point noise, especially the dust raised by the movement of the object, which causes the deformation and distortion of the scanned structure data; (2) Difficulty in point cloud motion detection. Point cloud data can only record the local coordinates of the hydraulic support at a certain time, but when the entire device moves, changes need to be found between multiple frames of data, which makes it difficult to measure the actual movement distance of the hydraulic support; (3) Difficulty in target detection. Using traditional point cloud registration or object detection methods, it is difficult to cope with the changes of the support itself, because the underground equipment environment is different, and the accessories on the equipment are also different, and the traditional pattern recognition method cannot confirm the specific pattern to be recognized, and the shape of the hydraulic support will change during the movement.

[0027] Currently, for real-time laser point cloud motion estimation, methods such as SLAM (Simultaneous localization and mapping), point cloud registration, and target recognition can be used. SLAM technology is not suitable for underground narrow environment, as it requires a large number of environmental feature points, which can easily fail in high noise conditions. The hydraulic support only changes in a local way, so the point cloud registration technology is not suitable for this global registration method, and the registration error correction will be restricted by a large number of fixed points. The target recognition method is mainly a deep learning method, but this method is not suitable for simple devices without GPU (Graphics Processing Unit) in the underground, and cannot be processed in real time. Traditional machine learning methods require artificial features, but such methods are sensitive to noise and are not suitable for target detection in high noise environments.

[0028] To this end, the embodiment provides a dynamic object moving distance measurement method based on laser point cloud. The method based on the embodiment can achieve accurate positioning of a moving target, thereby achieving accurate measurement of the moving distance of the dynamic object, and has high robustness. In a specific application, the embodiment first acquires current frame point cloud data of a moving target, performs registration on the current frame point cloud data and a preset target point cloud template sequence to obtain a best matching template. Then, the best matching template is moved to a target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data. Next, based on the reference point of the current frame point cloud data, it is determined whether the current frame point cloud data is end frame point cloud data. If the current frame point cloud data is end frame point cloud data, the moving distance of the moving target is obtained based on a reference point of initial frame point cloud data and the reference point of the current frame point cloud data.

[0029] The embodiment takes measuring the moving distance of a downhole hydraulic support as an example. The downhole hydraulic support mainly plays a role of moving the excavating equipment when excavating underground. When the excavating equipment needs to move, a moving signal is sent to the hydraulic support. After receiving the moving signal, the hydraulic support raises the height of the support and moves forward by a certain distance. After executing the moving instruction, the hydraulic support lowers the height to realize the fixation of the hydraulic support. Therefore, the purpose of the embodiment is to calculate the actual moving distance of the hydraulic support during each moving process.

[0030] In actual application, the downhole hydraulic support is divided into upper and lower parts. The lower part of the base moves first, and after the movement stops, the upper part of the support moves. The overall moving mode is to first raise, then move forward, and then lower to a reasonable height. In the embodiment, the laser radar equipment is installed at the top of the upper part of the support and scans the lower part of the support from above. Therefore, the field of view of the laser radar is the space range from the top support to the bottom base. The laser radar is installed at the top because the top position moves after the movement of the bottom base is completed. Therefore, the laser radar at the top can completely record the entire movement process of the bottom base during the stationary process, and the length of the measured distance depends only on the maximum movement distance of the bottom base. When the bottom base moves forward, the laser radar point cloud flow shows that the bottom base moves away from the coordinate origin of the laser radar. When the top moves forward, the laser radar point cloud flow shows that the bottom base moves close to the coordinate origin of the laser radar. Therefore, theoretically, the distance measured in real time is a gradual and progressive process of first becoming larger and then becoming smaller. The actual moving distance of the excavating equipment where the hydraulic support is located is the maximum value of the entire measurement process.

[0031] The dynamic object moving distance measurement method based on laser point cloud of the embodiment can be applied to a terminal. The terminal can be a computer, a smart television, or other intelligent devices. As shown in the embodiment, the method of the embodiment specifically includes the following steps: Figure 1 ​In step S100, current frame point cloud data of a moving target is acquired, the current frame point cloud data is registered with a preset target point cloud template sequence, and a best matching template is obtained.

[0032] The essence of the embodiment is to calculate the moving distance of a moving target before and after a period of time. Essentially, only the moving distance of the same point before and after movement of two frames needs to be found. Figure 2 As shown in the figure, the embodiment analyzes single-frame point cloud data, which is the frame-by-frame data of a real-time laser point cloud stream. The receiving frequency of single-frame data is 10 Hz, that is, new single-frame data is acquired every 0.1 seconds, so the processing time is only 0.1 seconds. The embodiment also needs to be assisted by a target point cloud template sequence, which is a set of artificial point cloud templates. Usually, the point cloud templates of each device are fixed. The moving target of the embodiment is a landmark component or landmark position on the scanned device, and has no specific size and range requirements, and weak semantic association. When the method of the embodiment is applied to the measurement of the moving distance of an underground hydraulic support, the moving target can be only the bottom of a specific support or the operating table part, etc. In actual application, the front position of the operating table of the underground hydraulic support is taken as the moving target, there is no other object above the moving target, and the moving target can move in real time with the device and is not easily affected by noise. The point cloud data obtained by scanning the front position of the operating table using a laser scanner is the point cloud data of the moving target, which can be recorded as a target point cloud template.

[0033] The target point cloud template sequence refers to the point cloud data of the moving target recorded multiple times in the entire movement cycle. When making the target point cloud template sequence, the embodiment scans the movement process of the moving target in the entire movement cycle to obtain the point cloud stream of the moving target. Then, the point cloud stream of the moving target is converted into single-frame point cloud data for storage, and the single-frame point cloud data is segmented using third-party software to obtain the point cloud data of the moving target of each frame. Because the point cloud data of the moving target of different frames appears at different positions due to movement during the movement of the moving target, different point cloud distributions in time and space are formed. Therefore, the embodiment can record the distribution of the point cloud data of the moving target of different frames in time and space to obtain the target point cloud template sequence. Because the movement mode of the device is fixed, matching the current frame point cloud data with the target point cloud template sequence can find the corresponding movement action and movement state of the current frame point cloud data. Specifically, Figure 3 The (a) figure in FIG. 1 is a target point cloud template sequence, which includes multiple target point cloud templates, and different target point cloud templates are distinguished by different colors. Figure 3The (b) figure in the figure is a single template. In step S100, the present embodiment registers the current frame point cloud data with each template in the target point cloud template sequence, thereby obtaining the best matching template, which is a coarse registration process.

[0034] The present embodiment obtains the point cloud data set of each template in the target point cloud template sequence. Then, the one-way Hausdorff distance between the current frame point cloud data and the point cloud data set of each template is calculated, and finally, the template corresponding to the minimum value of the one-way Hausdorff distance is taken as the best matching template. Specifically, let the point set of the single frame point cloud data at the current time be , that is, the point set of the current frame point cloud data is , the point cloud data set of the first template is , then the similarity between the two is calculated by using the one-way Hausdorff distance: .

[0035] Specifically, for each point in each template in the target point cloud template sequence , and the constituent point , is the position of a point in the template, there is . is a three-dimensional coordinate, , and . is the number of points in the template, which has no prior requirement and is consistent with the conventional number of points in single frame point cloud data. For single frame point cloud data , and the constituent point , is the position of a point in the single frame point cloud data, there is . is also a three-dimensional coordinate, , and . is the number of points in the single frame point cloud data, and when , the subsequent operation is performed, otherwise it is treated as a noise frame and directly skipped. The selected value depends on the stability of the laser radar, for example, the value is selected. When calculating the one-way Hausdorff distance, the one-way Hausdorff distance refers to the farthest distance from the template to the point cloud data in the template. The purpose of one-way calculation is to correctly find the proximity of the point cloud data of the moving target and the current frame point cloud data from the local point of view, otherwise from the point of view of the current frame point cloud data, there will always be a current frame point cloud data far away from the point cloud data of the moving target. At this time, the one-way Hausdorff distance is calculated as follows: .

[0036] For each template, a distance value corresponding to the template can be obtained. Thus, the Hausdorff distance calculated according to each template is The best matching template can be determined according to the Hausdorff distance calculated according to each template The formula is as follows: .

[0037] That is, the template corresponding to the minimum value of the one-way Hausdorff distance is taken as the best matching template in the embodiment. Thus, after the coarse registration process in step S100, the template Among all the templates, the template obtains the nearest distance, which indicates that the template is closest to the moving position of the current frame, that is, the current device is approximately moved to this action state.

[0038] It should be noted that the one-way Hausdorff distance in the embodiment is a point set distance measurement method, and other distances such as chamfer distance can also be used as a distance measurement method, and the embodiment is not limited thereto. In addition, the size and number of the target point cloud template in the embodiment are not limited, and the target point cloud template can be obtained by manual production or algorithm generation in specific application, and the properties of various templates do not affect the final effect.

[0039] Step S200: moving the best matching template to the target position corresponding to the current frame point cloud data to obtain the reference point of the current frame point cloud data.

[0040] After obtaining the best matching template, this embodiment performs a fine registration process. The principle of fine registration is to use the highest point as a guide, with the goal of further moving the best matching template to the potential target location. This movement is essentially a point move, moving the highest point of the best matching template to the highest point of a nearby potential target. This embodiment first determines the highest point of the best matching template, then determines the point in the current frame's point cloud data closest to the highest point. The best matching template and its corresponding highest point are then moved to the closest point, obtaining a point cloud dataset of the best matching template after the first move. Then, starting from the highest point of the best matching template after the first move, the highest point within a certain range of the current frame's point cloud data is determined. The best matching template after the first move is further moved to the highest point within the same range, obtaining a point cloud dataset of the best matching template after the second move. Next, the Hausdorff distances are calculated between the current frame's point cloud data and the point cloud dataset of the best matching template, the point cloud dataset of the best matching template after the first move, and the point cloud dataset of the best matching template after the second move. Finally, based on the minimum value of the Hausdorff distance, the reference point of the current frame point cloud data is obtained.

[0041] When applying it in specific Figure 4 As shown in , the precise registration in this embodiment mainly includes three steps: (1) calculating the highest point of the best matching template ( Figure 4 dotted box in the upper left corner). , then calculate the point closest to the highest point in the current frame point cloud data based on the highest point of the best matching template , ( Figure 4 The second sequence arrow from left to right points to the position). Then, move the highest point and the best matching template to the nearest point. , the point cloud dataset of the best matching template after the first movement is (2)The highest point of the best matching template after the first move As the starting point, continue to search for a certain range in the current frame point cloud data The highest point within , the best matching template after the first move is moved from Move to , the point cloud dataset of the best matching template after the second move is .in, and Respectively represent and exist The values ​​on the axis ( Component value). (3) Calculate the number of point clouds in the current frame and the best matching template obtained by rough registration respectively , the point set after the first step of the best matching template movement , the point set after the second step of the best matching template movement The Hausdorff distances between 、 、 , and retain the minimum value in Hausdorff distance: .

[0042] in, Current frame The Hausdorff distance value after the above-mentioned fine registration process. The reference point corresponding to the minimum value in the Hausdorff distance is the reference point of the current frame point cloud data. , specifically: .

[0043] like Figure 4 As shown in the figure, the red dotted box is the point cloud shape of the best matching template, the green solid box is the point cloud shape near the current frame point cloud data, the orange point is the reference point for mobile calculation, and the order of fine registration is from left to right. Calculation reference point This is the highest point and represents the moving position of the moving object in the current frame. This fine registration often achieves a local optimum. Even when the local high point contains significant noise, the coarse registration accuracy is maintained, ensuring no degradation. This approach ensures real-time performance. The reference point only needs to be moved twice, and in the absence of noise, the optimal match can be found.

[0044] like Figure 5 As shown, Figure 5 A precise registration example is given. Figure 5 The gray points in the figure are local points of the current frame point cloud data, the blue points are the point cloud of the best matching template obtained after coarse registration, and the blue ball is the highest point of the best matching template; the red points are the point cloud of the best matching template after the first movement, and the red ball is the reference point at this time; the green points are the point cloud of the best matching template after the second movement, and the green ball is the reference point at this time.

[0045] Step S300: Based on the reference point of the current frame point cloud data, determine whether the current frame point cloud data is the end frame point cloud data; if the current frame point cloud data is the end frame point cloud data, obtain the moving distance of the moving target based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data.

[0046] After determining the reference point of the current frame point cloud data, the embodiment calculates the Euclidean distance between the reference point of the current frame point cloud data and the reference point of the initial frame point cloud data, and then determines whether the Euclidean distance meets the motion end condition based on the Euclidean distance and the distance mean calculated according to the time window in advance. If the motion end condition is met, the current frame point cloud data is determined as the end frame point cloud data. The determination of the end frame point cloud data can be determined by an algorithm or by the signal input of the hydraulic support itself.

[0047] Specifically, in the embodiment, it is determined whether the current frame point cloud data is the end frame point cloud data, that is, whether the Euclidean distance between the reference point of the current frame point cloud data and the reference point of the initial frame point cloud data is the maximum value in the device movement process. Let the reference point of the initial frame point cloud data before movement be , and the reference point of the current frame point cloud data during movement be , the Euclidean distance between the two frames of reference points is . .

[0048] The values of the device motion initial frame and the end frame can be obtained by signals, that is, the device receives the instruction of device movement or stop at a certain time. However, the embodiment mainly adopts the way of autonomously estimating the motion period, and the motion initial time is recorded as , and the calculated reference point after coarse registration and fine registration is . Therefore, there is a corresponding value for each time, and the best value is: .

[0049] The embodiment can convert the problem into how to calculate the value of the motion end time , that is, how to determine that the device has completed the movement. Although the symbol of ending the movement is to obtain the maximum value, it is obvious that even if is obtained, it cannot be immediately determined that the movement is ended, because there may be noise and registration errors, etc., which cause local jumps. Therefore, the embodiment introduces a time window mean function as a judgment standard, that is, for a time window , is the distance mean in the motion period: .

[0050] The distance mean represents the mean value of the measured movement distance in a certain time, and usually . Obviously, when the device is stationary, there will be . Therefore, the condition for determining the end of the movement is: .

[0051] in, Is the distance threshold, usually 0.05. Based on this, when the device's moving distance is measured within the time period Within the range, the difference between the current measured distance value and the average distance is less than This method can autonomously determine the start and stop time of the device without any external signal input as a reference. and middle and The size of , and also obtained the best Euclidean distance value. The corresponding best matching template , and determine the current frame point cloud data The end frame point cloud data. That is, the moving distance of the moving target.

[0052] Furthermore, if the fine registration process is correct enough, it is possible to directly As the best result. However, in order to ensure real-time performance, the precise registration only uses three registration steps, which may result in a certain error. In order to obtain an accurate solution, this implementation analyzes all points in the current frame point cloud data point by point, such as Figure 2 As shown in , the optimal reference point for the point cloud data of the current frame is obtained, and thus, the most accurate movement distance of the moving target is obtained based on the optimal reference point. In addition, this embodiment performs point-by-point registration on the initial frame and the final frame, thereby obtaining accurate calculation reference points. The Euclidean distance between the two calculation reference points is then calculated to obtain the final movement distance of the device.

[0053] In summary, this embodiment can form a target point cloud template sequence by recording all possible moving positions, and then compare each frame of point cloud data with the target point cloud template sequence to find the best matching template, and then move the best matching template for precise alignment to achieve precise positioning of the moving position. In addition, this embodiment calculates the distance between the target point cloud template and the entire single-frame point cloud data through a one-way Hausdorff distance, and at the same time uses the highest point traction to achieve the movement of the best matching template to the corresponding target position in the single-frame point cloud data. Finally, during the calculation, the precise movement value of the best matching template is solved by traversing all points. In addition, this embodiment calculates the rough movement results of each frame in real time, and then intelligently selects the initial frame and the end frame of the movement through the distance average, and finally calculates the distance between the initial frame and the end frame.

[0054] The method of the present application has been fully tested and verified, and has been run on a large number of physical machines in the well, and accurate results have been obtained, which are consistent with the results of manual measurement, with an error of centimeter level. The experiment and verification of the present application are carried out in the hydraulic support equipment working in the real mine, and the laser radar is used and installed on the wall of the hydraulic support of the equipment, that is, the top of the equipment. When the equipment moves, the equipment below the hydraulic support moves first, so that the laser radar can record the whole movement process, and the point cloud data of the moving target on the laser radar is away from the origin of the local coordinate of the laser radar. When the equipment below stops moving, the equipment above the support starts to move forward and is retracted to match the position of the equipment below, at which time the point cloud data of the moving target is close to the origin of the local coordinate of the laser radar. Through the method of the present application, the moving distance of the equipment in the well is accurately measured, and the measurement response time is fast, the accuracy is high, and the error is centimeter level, effectively replacing the traditional manual measurement method.

[0055] The method of the present application is based on real-time laser point cloud, and the obtained moving distance measurement result is also real-time calculation, so the result can be obtained immediately after the hydraulic support moves, and the application value is high. The present application can maximize the matching accuracy to the data point scale through coarse matching, fine matching and point-by-point analysis processes. The method of the present application utilizes the finiteness of the motion space, so that the final result is within a reasonable range. At the same time, the maximum moving distance is obtained by the distance mean value, which has high robustness.

[0056] Based on the above embodiment, the present application also provides a dynamic object moving distance measurement system based on laser point cloud, which can be used to realize the steps in the above method embodiment. As shown in Figure 6 The coarse registration module 10 is used to obtain the current frame point cloud data of the moving target, and register the current frame point cloud data with a preset target point cloud template sequence to obtain a best matching template. The fine registration module 20 is used to move the best matching template to a target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data. The distance determination module 30 is used to judge whether the current frame point cloud data is end frame point cloud data based on the reference point of the current frame point cloud data, and if the current frame point cloud data is end frame point cloud data, obtain the moving distance of the moving target based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data.

[0057] The principles of each device or module in the dynamic object moving distance measurement system based on laser point cloud of the present embodiment are the same as the working principles of each step in the above method embodiment, which will not be repeated here.

[0058] Each module in the aforementioned laser point cloud-based dynamic object distance measurement system can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of the processor in the terminal in hardware form, or can be stored in the terminal's memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0059] Based on the above embodiment, the present invention further provides a terminal, the principle block diagram of the terminal can be as follows: Figure 7 The terminal may include one or more processors 100 ( Figure 7 Only one is shown in the figure), memory 101, and computer program 102 stored in memory 101 and executable on one or more processors 100. For example, a program for measuring the distance of a moving object based on a laser point cloud. When one or more processors 100 execute computer program 102, each step of an embodiment of a method for measuring the distance of a moving object based on a laser point cloud can be implemented. Alternatively, when one or more processors 100 execute computer program 102, each module / unit in an embodiment of a system for measuring the distance of a moving object based on a laser point cloud can be implemented, without limitation herein.

[0060] In one embodiment, the processor 100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0061] In one embodiment, memory 101 may be an internal storage unit of an electronic device, such as a hard drive or memory. Memory 101 may also be an external storage device of the electronic device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash memory card, etc. Furthermore, memory 101 may include both an internal storage unit of the electronic device and an external storage device. Memory 101 is used to store computer programs and other programs and data required by the terminal. Memory 101 may also be used to temporarily store data that has been output or is about to be output.

[0062] Those skilled in the art can understand that, Figure 7 The principle block diagram shown in the above-mentioned embodiments is only a block diagram of part of the structure related to the technical scheme of the present application, and does not constitute a limitation on the terminal to which the technical scheme of the present application is applied. The specific terminal can include more or less components than those shown in the diagram, or combine certain components, or have a different arrangement of components.

[0063] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing relevant hardware. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiments. Any reference to memory, storage, operating database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM) and the like.

[0064] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications 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 application.

Claims

1. A method for measuring the moving distance of a dynamic object based on laser point cloud, characterized in that: The method comprises: Acquire the current frame point cloud data of the moving target, and align the current frame point cloud data with a preset target point cloud template sequence to obtain the best matching template; Moving the best matching template to the target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data; Based on the reference point of the current frame point cloud data, determine whether the current frame point cloud data is the end frame point cloud data. If the current frame point cloud data is the end frame point cloud data, then based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data, obtain the moving distance of the moving target.

2. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, characterized in that: The moving target is a landmark component or landmark position on the scanned device. When the method is applied to measure the moving distance of an underground hydraulic support, the moving target is the front position of the working table of the underground hydraulic support.

3. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, characterized in that: The method of preparing the target point cloud template sequence includes: Scanning the movement process of the moving target during the entire movement cycle to obtain a point cloud stream of the moving target; Converting the point cloud stream of the moving target into single-frame point cloud data for storage, and segmenting the single-frame point cloud data to obtain point cloud data of the moving target in each frame; The target point cloud template sequence is obtained according to the temporal and spatial distribution of the point cloud data of the moving target in different frames.

4. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, wherein: Registering the current frame point cloud data with a preset target point cloud template sequence to obtain the best matching template includes: Acquire a point cloud dataset of each template in the target point cloud template sequence; Calculate the one-way Hausdorff distance between the current frame point cloud data and the point cloud dataset of each template; The template corresponding to the minimum value of the one-way Hausdorff distance is used as the best matching template.

5. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, characterized in that: Moving the best matching template to a target position corresponding to the current frame point cloud data to obtain a reference point of the current frame point cloud data includes: Determine the highest point of the best matching template, determine the point in the current frame point cloud data that is closest to the highest point, and move the best matching template and the corresponding highest point to the point that is closest to the highest point, to obtain a point cloud dataset of the best matching template after the first movement; Taking the highest point of the best matching template after the first movement as the starting point, determining the highest point within a certain range of the point cloud data of the current frame, and continuing to move the best matching template after the first movement to the highest point within the certain range, to obtain the point cloud dataset of the best matching template after the second movement; Calculating the Hausdorff distances between the current frame point cloud data and the point cloud dataset of the best matching template, the point cloud dataset of the best matching template after the first movement, and the point cloud dataset of the best matching template after the second movement respectively; Based on the minimum value of the Hausdorff distance, a reference point of the current frame point cloud data is obtained.

6. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, characterized in that: Determining whether the current frame point cloud data is end frame point cloud data based on a reference point of the current frame point cloud data includes: Calculating the Euclidean distance between the reference point of the current frame point cloud data and the reference point of the initial frame point cloud data; Based on the Euclidean distance and a distance mean pre-calculated according to the time window, determining whether the Euclidean distance meets the movement end condition; If the motion end condition is met, the current frame point cloud data is determined to be the end frame point cloud data.

7. The method for measuring the moving distance of a dynamic object based on laser point cloud according to claim 1, characterized in that: The method further comprises: All points in the current frame point cloud data are analyzed point by point to obtain the best reference point of the current frame point cloud data, and the moving distance of the moving target is obtained according to the best reference point.

8. A dynamic object moving distance measurement system based on laser point cloud, characterized in that: The system comprises: A coarse registration module is used to obtain the current frame point cloud data of the moving target, and to register the current frame point cloud data with a preset target point cloud template sequence to obtain the best matching template; A fine registration module is used to move the best matching template to the target position corresponding to the current frame point cloud data to obtain the reference point of the current frame point cloud data; A distance determination module is used to determine whether the current frame point cloud data is the end frame point cloud data based on the reference point of the current frame point cloud data. If the current frame point cloud data is the end frame point cloud data, the moving distance of the moving target is obtained based on the reference point of the initial frame point cloud data and the reference point of the current frame point cloud data.

9. A terminal, characterized in that: The terminal includes a memory, a processor, and a dynamic object movement distance measurement program based on laser point cloud stored in the memory and runnable on the processor. When the processor executes the dynamic object movement distance measurement program based on laser point cloud, the steps of the dynamic object movement distance measurement method based on laser point cloud as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program for measuring the moving distance of a dynamic object based on a laser point cloud, and the program for measuring the moving distance of a dynamic object based on a laser point cloud implements the steps of the method for measuring the moving distance of a dynamic object based on a laser point cloud as described in any one of claims 1 to 7 on the computer-readable storage medium.

Images