A method and system for measuring the volume of automobile trunk based on laser scanning

Abstract

The present application relates to a method and system for measuring the volume of a car trunk based on laser scanning, and relates to the field of volume measurement technology. The method comprises the following steps: using a preset laser radar to scan the interior of a car trunk to obtain original three-dimensional point cloud data; performing noise reduction and simplification processing on the original three-dimensional point cloud data to obtain simplified point cloud data; and calculating the volume of the car trunk based on the simplified point cloud data. The present application is based on scanning the interior of the trunk by a laser radar, and calculating the trunk volume based on the point cloud data. Compared with traditional technical means, the present application has higher accuracy and reliability, higher calculation efficiency, and lower cost.

Description

Technical Field

[0001] The present application relates to the technical field of volume measurement, and in particular to a method and system for measuring the volume of a car trunk based on laser scanning. Background Art

[0002] As the popularity of automobiles increases, people have more and more stringent requirements on automobile functions. When measuring automobile parameters, the trunk volume is a very important indicator, especially for practical models such as SUVs and family station wagons. Being able to quickly and accurately measure the trunk volume is particularly important for automobile manufacturers and testing agencies.

[0003] There are currently two common methods for measuring the volume of a car trunk. One is to place reference blocks of standard volume in the trunk in sequence until the trunk is completely filled, then take them out and count the number, and finally calculate the total volume of the reference blocks as the trunk volume. However, this method requires a lot of manpower and material resources, and is inefficient. In addition, due to the irregular shape of the trunk space, it is difficult for the reference blocks to completely fill the entire trunk, and the measurement result is small, so the measurement accuracy is also low. The second method is to inflate the flexible air bag multiple times through the inflation mechanism, and set a pressure sensor and set a threshold in the irregular area of ​​the trunk until the air bag triggers the pressure threshold and fills the trunk. The trunk volume is read out according to the volume matrix table corresponding to the pre-stored inflation number. This method has a certain improvement in measurement efficiency compared to the former, but the placement of the pressure sensor and the setting of the pressure threshold are relatively complicated, and the air pressure is greatly affected by the environment, which makes the measurement result repeatability low.

[0004] Therefore, a new automobile trunk volume measurement technology is now needed to meet current usage needs. Summary of the invention

[0005] The present application provides a method and system for measuring the volume of a car trunk based on laser scanning. The method and system scan the interior of the trunk based on a laser radar and calculate the volume of the trunk according to point cloud data. Compared with traditional technical means, the method and system have higher accuracy and reliability, higher measurement efficiency, and lower cost.

[0006] In a first aspect, the present application provides a method for measuring the volume of a car trunk based on laser scanning, the method comprising the following steps:

[0007] Use the preset LiDAR to scan the interior of the car trunk to obtain original 3D point cloud data;

[0008] Performing noise reduction and simplification processing on the original three-dimensional point cloud data to obtain simplified point cloud data;

[0009] Based on the simplified point cloud data, the volume of the car trunk is calculated.

[0010] Specifically, the original three-dimensional point cloud data is subjected to noise reduction and simplification processing to obtain simplified point cloud data, including the following steps:

[0011] Performing large-scale noise removal on the original three-dimensional point cloud data to obtain coarse denoised point cloud data;

[0012] Calculating curvature based on the coarse denoised point cloud data, and partitioning regions according to the curvature into flat region points and feature region points;

[0013] Perform fine denoising on flat area points and feature area points respectively to obtain denoised point cloud data;

[0014] Determine edges based on the denoised point cloud data, and evenly divide the data to obtain a plurality of point cloud blocks;

[0015] Calculating the average curvature of the point cloud block, and performing point cloud feature partitioning to divide it into a flat area and a feature area;

[0016] The flattened area and the characteristic area are streamlined to obtain streamlined point cloud data.

[0017] Specifically, based on the simplified point cloud data, calculating the volume of the car trunk includes the following steps:

[0018] A series of point cloud slices are obtained by cutting the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes;

[0019] Calculate the slice area S according to the contour boundary of each point cloud slice i , and then calculate the trunk volume.

[0020] Specifically, the trunk volume calculation formula is:

[0021]

[0022] Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

[0023] Specifically, the single-line point cloud data set corresponding to each frame of laser radar data is P = {p1, p2, p3...p n}, n is the number of midpoints of the line;

[0024] The original three-dimensional point cloud data is p(x, y, z);

[0025]

[0026] α=ωt;

[0027] d is the distance between the laser radar and point p, β is the pitch angle between the laser radar and point p;

[0028] ω is the angular velocity of the laser radar during scanning;

[0029] The time corresponding to the data of the laser radar collection point p is the tth second, and the relative angle between the radar direction and the initial position at this time is α.

[0030] Specifically, performing large-scale noise removal on the plurality of original three-dimensional point cloud data to obtain coarse denoised point cloud data includes the following steps:

[0031] Based on the original 3D point cloud data, calculate the corresponding local density ρ i The distance δ from the nearest high density point i ;

[0032] P i and δ i Arrange in descending order, based on the preset noise point judgment threshold s, i In the first s% and ρ i The original 3D point cloud data located in the last s% are treated as scattered noise points and removed;

[0033] Clustering is performed based on the cluster center weights, and clusters whose number of points in the clusters after clustering is less than the preset threshold are removed as cluster noise points to obtain coarse denoised point cloud data.

[0034] Specifically, the local density calculation formula of the local density is:

[0035] ρ i =∑ j≠i x(d ij -d c );

[0036] The calculation formula of the distance between the nearest high-density point and the nearest high-density point is:

[0037]

[0038] in,

[0039] d ij For point p j To point p i The distance, d c is the cutoff distance.

[0040] Specifically, the cluster center weight calculation formula of the cluster center weight is:

[0041] γ i =ρ i *δ i .

[0042] In a second aspect, the present application provides a vehicle trunk volume measurement system based on laser scanning, the system comprising:

[0043] LiDAR, which is used to scan the interior of the car trunk to obtain raw 3D point cloud data;

[0044] A data reduction element is used to reduce noise and reduce the original three-dimensional point cloud data to obtain reduced point cloud data;

[0045] A volume calculation unit is used to calculate the volume of the car trunk based on the simplified point cloud data.

[0046] Specifically, the system further includes:

[0047] A rotating platform is used to control the laser radar to rotate at a preset angular velocity.

[0048] Specifically, data reduction elements include:

[0049] A coarse denoising module, which is used to remove large-scale noise from the original three-dimensional point cloud data to obtain coarse denoised point cloud data;

[0050] A coarse denoising area point division module, which is used to calculate the curvature based on the coarse denoising point cloud data, and divide the area into flat area points and feature area points according to the curvature;

[0051] A fine denoising module is used to perform fine denoising on flat area points and feature area points respectively to obtain denoised point cloud data;

[0052] A point cloud block division module, which is used to determine the edge based on the denoised point cloud data and divide it evenly to obtain multiple point cloud blocks;

[0053] A fine denoising area division module, which is used to calculate the average curvature of the point cloud block and perform point cloud feature partitioning into a flat area and a feature area;

[0054] A fine denoising and simplification module is used to perform simplification processing on the flattened area and the feature area to obtain simplified point cloud data.

[0055] Specifically, the volume calculation unit includes:

[0056] A point cloud slicing module, which is used to cut the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes to obtain a series of point cloud slices;

[0057] Volume calculation module, which is used to calculate the slice area S according to the contour boundary of each point cloud slice i , and then calculate the trunk volume.

[0058] Specifically, the trunk volume calculation formula is:

[0059]

[0060] Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

[0061] Specifically, the single-line point cloud data set corresponding to each frame of laser radar data is P = {p1, p2, p3...p n}, n is the number of midpoints of the line;

[0062] The original three-dimensional point cloud data is p(x, y, z);

[0063]

[0064] α=ωt;

[0065] d is the distance between the laser radar and point p, β is the pitch angle between the laser radar and point p;

[0066] ω is the angular velocity of the laser radar during scanning;

[0067] The time corresponding to the data of the laser radar collection point p is the tth second, and the relative angle between the radar direction and the initial position at this time is α.

[0068] Specifically, the coarse denoising module is also used to calculate the corresponding local density ρ based on the original 3D point cloud data. i The distance δ from the nearest high density point i ;

[0069] The coarse denoising module is also used to i and δ i Arrange in descending order, based on the preset noise point judgment threshold s, i In the first s% and ρ i The original 3D point cloud data located in the last s% are treated as scattered noise points and removed;

[0070] The coarse denoising module is also used to perform clustering based on the cluster center weights, and remove clusters whose number of points in the clusters after clustering is less than a preset threshold as cluster noise points to obtain coarse denoised point cloud data.

[0071] Specifically, the local density calculation formula of the local density is:

[0072] ρ i =∑ j≠i x(d ij -d c );

[0073] The calculation formula of the distance between the nearest high-density point and the nearest high-density point is:

[0074]

[0075] in,

[0076] d ij For point p j To point p i The distance, d c is the cutoff distance.

[0077] Specifically, the cluster center weight calculation formula of the cluster center weight is:

[0078] γ i =ρ i *δ i .

[0079] The beneficial effects of the technical solution provided by this application include:

[0080] This application is based on laser radar scanning the inside of the trunk and calculates the trunk volume based on point cloud data. Compared with traditional technical means, it has higher accuracy and reliability, higher measurement efficiency and lower cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Terminology explanation:

[0082] PCA: Principal Components Analysis, principal component analysis technology;

[0083] DPC: Desk Personal Computer, desktop personal computer.

[0084] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0085] Figure 1 A flowchart of the steps of the automobile trunk volume measurement method based on laser scanning provided in an embodiment of the present application;

[0086] Figure 2 This is a flowchart of step S2 in the method for measuring the volume of a car trunk based on laser scanning provided in an embodiment of the present application;

[0087] Figure 3This is a structural block diagram of a car trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0088] Figure 4 This is a schematic diagram of the structure of a car trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0089] Figure 5 A schematic diagram of the laser radar measurement principle in the automobile trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0090] Figure 6 This is a working principle diagram of the laser radar and the rotating table in the automobile trunk volume measurement system based on laser scanning provided in the embodiment of the present application;

[0091] Figure 7 A schematic diagram of the principle of acquiring three-dimensional point cloud data in a car trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0092] Figure 8 A simplified flowchart of point cloud data denoising in a car trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0093] Fig. 9 A volume calculation principle diagram of a car trunk volume measurement system based on laser scanning provided in an embodiment of the present application;

[0094] Fig.10 This is a flow chart of automobile trunk volume measurement in the automobile trunk volume measurement system based on laser scanning provided in an embodiment of the present application. DETAILED DESCRIPTION

[0095] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0096] The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0097] The embodiment of the present application provides a method and system for measuring the volume of a car trunk based on laser scanning. The method and system calculate the volume of the trunk based on point cloud data based on laser radar scanning of the inside of the trunk. Compared with traditional technical means, the method has higher accuracy and reliability, higher measurement efficiency, and lower cost.

[0098] In order to achieve the above technical effects, the overall idea of ​​this application is as follows:

[0099] A method for measuring the volume of a car trunk based on laser scanning, the method comprising the following steps:

[0100] S1. Use the preset laser radar to scan the interior of the car trunk to obtain original three-dimensional point cloud data;

[0101] S2, performing noise reduction and simplification processing on the original three-dimensional point cloud data to obtain simplified point cloud data;

[0102] S3. Based on the simplified point cloud data, calculate and obtain the volume of the car trunk.

[0103] The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0104] First, see Figures 1-2 As shown, the embodiment of the present application provides a method for measuring the volume of a car trunk based on laser scanning, and the method comprises the following steps:

[0105] S1. Use the preset laser radar to scan the interior of the car trunk to obtain original three-dimensional point cloud data;

[0106] S2, performing noise reduction and simplification processing on the original three-dimensional point cloud data to obtain simplified point cloud data;

[0107] S3. Based on the simplified point cloud data, calculate and obtain the volume of the car trunk.

[0108] The technical solution of the embodiment of the present application is based on scanning the inside of the trunk with a laser radar and calculating the trunk volume according to point cloud data. Compared with traditional technical means, it has higher accuracy and reliability, higher measurement efficiency, and lower cost.

[0109] Specifically, the original three-dimensional point cloud data is subjected to noise reduction and simplification processing to obtain the simplified point cloud data, including the following steps:

[0110] S20, performing large-scale noise removal on the original three-dimensional point cloud data to obtain coarse denoised point cloud data;

[0111] S21, calculating curvature based on the coarse denoised point cloud data, and partitioning the region according to the curvature into flat region points and feature region points;

[0112] S22, performing fine denoising processing on the flat area points and the feature area points respectively to obtain denoised point cloud data;

[0113] S23, determining edges based on the denoised point cloud data, and evenly dividing the data to obtain a plurality of point cloud blocks;

[0114] S24, calculating the average curvature of the point cloud block, and performing feature partitioning of the point cloud into a flat area and a feature area;

[0115] S25. Simplify the flattened area and the feature area to obtain simplified point cloud data.

[0116] It should be noted that when calculating the point cloud curvature, the following methods can be used:

[0117] The common principal component analysis (PCA) method is used to calculate the curvature of the rough denoised point cloud data:

[0118]

[0119]

[0120] Where: C is the covariance matrix of point p, R is the curvature of point p, p j are the k-neighborhood points of point p, and λ1, λ2, and λ3 are the three eigenvalues ​​obtained by eigendecomposition of matrix C.

[0121] In addition, in the feature partitioning operation based on curvature, assuming that a point p in the coarse denoising data i The curvature is R i , the average curvature in its neighborhood is Then define point p i The local feature judgment weight in its k-neighborhood is:

[0122]

[0123] By setting the threshold ω m Perform feature point judgment, if point p i The local feature weight ω(p i , k)>ω m , then we think that p i is a feature point, otherwise it is a flat area point.

[0124] Specifically, in the fine denoising process for flat area points and feature area points, the following two denoising methods can be used:

[0125] The first method is to use orthogonal global least squares method for denoising:

[0126] First, build a fitting plane model:

[0127]

[0128] in:

[0129] a, b, c are the fitting plane parameters, are the average coordinates of the neighborhood points.

[0130] Then, calculate the distance d from the point to the fitting plane i :

[0131]

[0132] Finally, set the distance threshold d max , when d i >d max , the point is deleted as a noise point, otherwise it is retained.

[0133] The second method is to use adaptive bilateral filtering for denoising:

[0134] p′=p+α*n

[0135] Where: α is the bilateral filter factor, and the original coordinate p is adjusted according to α and the normal vector n to obtain the new coordinate p'.

[0136] In addition, the calculation formula of α is as follows:

[0137]

[0138]

[0139]

[0140] Where: k is the number of points in the neighborhood of sampling point p, p i is the neighborhood point, ω c and ω s They are the smoothing filter weight function and the feature preservation weight function, and the parameter σ that controls the size of the two functions c and σ s They correspond to the smoothness and feature preservation of bilateral filtering respectively.

[0141] Usually, σ c Take the neighborhood radius, when σ c After determining, σ s The value of has a great influence on the smoothing effect. s The larger it is, the larger the normal correction distance of p is and the greater the degree of smoothing is.

[0142] σ c =max||pp i ||;

[0143] σ s =t*σ c ;

[0144]

[0145] Among them: t is the feature preservation weight, R is the curvature of point p, is the average curvature of the p-neighborhood.

[0146] After performing fine denoising on different areas of the coarse denoised point cloud data using different methods, the denoised point cloud data is finally obtained.

[0147] In the work of edge determination of point cloud data, the specific situation is as follows:

[0148] After completing the point cloud data denoising, the boundary points of the denoised point cloud data are judged and retained to determine the point p i Taking whether it is a boundary point as an example, the specific implementation steps are as follows:

[0149] Using point p i The k-neighborhood of constructs the fitting plane;

[0150] The p i Project each point in its neighborhood onto the fitting plane and connect p i And each neighborhood point obtains a vector set;

[0151] Select any vector in the vector set as the reference vector, and calculate the angle θ between each vector and the reference vector;

[0152] Set the angle threshold θ s , generally take 2 / 3π, and set the maximum angle θ max With threshold θ s Compare, if θ max >θ s , then we consider point p i is the boundary point.

[0153] In the work of dividing point cloud data into blocks, the specific situation is as follows:

[0154] The denoised point cloud data is grouped by k-means clustering. The specific steps are as follows:

[0155] Determine the corresponding number of clusters k according to the reduction rate g and the amount of point data N:

[0156] k = N (1-g);

[0157] Randomly select k points from the point cloud data as the initial cluster centers, calculate the distances from all remaining points to each cluster center, and assign these points to the clusters where the nearest cluster centers are located according to the distances, thus completing the initial clustering;

[0158] Then calculate the center point of each cluster, take the point in the cluster that is closest to the center point as the new cluster center and cluster again, repeat the above steps until the cluster center no longer changes, and the clustering is completed.

[0159] In the simplification work of different curvature areas, the specific situation is as follows:

[0160] The group with higher average curvature is subdivided again by k-means clustering, and the group with lower curvature remains unchanged. Finally, all points in all groups except the cluster center and boundary points are removed to achieve the effect of streamlining the point cloud data, and finally the streamlined point cloud data is obtained.

[0161] Specifically, based on the simplified point cloud data, calculating the volume of the car trunk includes the following steps:

[0162] S30, cutting the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes to obtain a series of point cloud slices;

[0163] S31, calculate the slice area S according to the contour boundary of each point cloud slice i , and then calculate the trunk volume.

[0164] Specifically, the trunk volume calculation formula is:

[0165]

[0166] Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

[0167] Specifically, the single-line point cloud data set corresponding to each frame of laser radar data is P = {p1, p2, p3...p n}, n is the number of midpoints of the line;

[0168] The original three-dimensional point cloud data is p(x, y, z);

[0169]

[0170] α=ωt;

[0171] d is the distance between the laser radar and point p, β is the pitch angle between the laser radar and point p;

[0172] ω is the angular velocity of the laser radar during scanning;

[0173] The time corresponding to the data of the laser radar collection point p is the tth second, and the relative angle between the radar direction and the initial position at this time is α.

[0174] Specifically, performing large-scale noise removal on the plurality of original three-dimensional point cloud data to obtain coarse denoised point cloud data includes the following steps:

[0175] S200, based on the original 3D point cloud data, calculate the corresponding local density ρ i The distance δ from the nearest high density pointi ;

[0176] S201, for ρ i and δ i Arrange in descending order, based on the preset noise point judgment threshold s, i In the first s% and ρ i The original 3D point cloud data located in the last s% are treated as scattered noise points and removed;

[0177] S202, clustering is performed based on the cluster center weights, and clusters whose number of points in the clusters after clustering is less than a preset threshold are removed as cluster noise points to obtain coarse denoised point cloud data.

[0178] Specifically, the local density calculation formula of the local density is:

[0179] ρ i =∑ j≠i x(d ij -d c );

[0180] The calculation formula of the distance between the nearest high-density point and the nearest high-density point is:

[0181]

[0182] in,

[0183] d ij For point p j To point p i The distance, d c is the cutoff distance.

[0184] Specifically, the cluster center weight calculation formula of the cluster center weight is:

[0185] γ i =ρ i *δ i .

[0186] Based on the technical solution of the embodiment of the present application, the automobile trunk volume measurement method based on laser scanning includes the following steps when actually implemented:

[0187] The first step is to set the working parameters of the laser radar and the rotating stage:

[0188] Set the rotation angular velocity of the rotating table to ω and the laser radar sampling frequency to fs. The laser radar and the rotating table start working at the same time from the starting position until the rotating table completes a circle and returns to the starting position.

[0189] The second step is to open the trunk lid of the car, place the measuring device near the center of the bottom of the trunk, raise the bracket so that the rotating table maintains a certain height relative to the bottom of the trunk to prevent the laser radar from being too close to the inner surface of the trunk and causing measurement errors.

[0190] The third step is to turn on the measuring device and check whether the equipment is working properly.

[0191] The fourth step is to confirm that the equipment is ready, close the car trunk lid, and after the host computer issues a start command, the device enters the working state. The wireless transmission module sends control commands downward and uploads the collected data. The control module controls the radar and the turntable to work according to the preset parameters to ensure that the lidar collected data corresponds to the turntable angle data.

[0192] The fifth step is to process the collected data through the data processing module of the host computer. It is known that the data collected by the laser radar per frame is a single-line point cloud data set P = {p1, p2, p3...p n}, n is the number of points in the line, and the single point p n (d n ,β n ), d n is the laser radar distance point p n The distance, β n is the laser radar and point p n The pitch angle.

[0193] Assuming that the time corresponding to a frame of data collected by the radar is the tth second, and t is an integer multiple of 1 / fs, the relative angle between the radar direction and the initial position at this time is α, and the calculation formula of α is:

[0194] α=ωt

[0195] In addition, after obtaining α, combined with p n The known β n ,d n The point cloud coordinate data of each point can be obtained, that is, point p (x, y, z), and the calculation formula is:

[0196]

[0197] In the sixth step, after obtaining the trunk point cloud coordinate data through coordinate transformation, the point cloud data is denoised and de-noised using the denoising algorithm based on the improved DPC and feature partitioning, and then the curvature-based bounding box method is used to simplify the data.

[0198] In the seventh step, after the point cloud data coordinates are converted, denoised, and simplified, a suitable method is used to calculate the point cloud volume, i.e., the volume of the car trunk.

[0199] Second, see Figures 3 to 10As shown, the embodiment of the present application provides a system for measuring the volume of a car trunk based on laser scanning on the basis of the technology of the car trunk volume measurement method based on laser scanning mentioned in the first aspect, and the system includes:

[0200] LiDAR, which is used to scan the interior of the car trunk to obtain raw 3D point cloud data;

[0201] A data reduction element is used to reduce noise and reduce the original three-dimensional point cloud data to obtain reduced point cloud data;

[0202] A volume calculation unit is used to calculate the volume of the car trunk based on the simplified point cloud data.

[0203] Specifically, the system also includes:

[0204] A rotating platform is used to control the laser radar to rotate at a preset angular velocity.

[0205] Specifically, data reduction elements include:

[0206] A coarse denoising module, which is used to remove large-scale noise from the original three-dimensional point cloud data to obtain coarse denoised point cloud data;

[0207] A coarse denoising area point division module, which is used to calculate the curvature based on the coarse denoising point cloud data, and divide the area into flat area points and feature area points according to the curvature;

[0208] A fine denoising module is used to perform fine denoising on flat area points and feature area points respectively to obtain denoised point cloud data;

[0209] A point cloud block division module, which is used to determine the edge based on the denoised point cloud data and divide it evenly to obtain multiple point cloud blocks;

[0210] A fine denoising area division module, which is used to calculate the average curvature of the point cloud block and perform point cloud feature partitioning into a flat area and a feature area;

[0211] A fine denoising and simplification module is used to perform simplification processing on the flattened area and the feature area to obtain simplified point cloud data.

[0212] Specifically, the volume calculation unit includes:

[0213] A point cloud slicing module, which is used to cut the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes to obtain a series of point cloud slices;

[0214] Volume calculation module, which is used to calculate the slice area S according to the contour boundary of each point cloud slice i, and then calculate the trunk volume.

[0215] Specifically, the trunk volume calculation formula is:

[0216]

[0217] Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

[0218] Specifically, the single-line point cloud data set corresponding to each frame of laser radar data is P = {p1, p2, p3...p n}, n is the number of midpoints of the line;

[0219] The original three-dimensional point cloud data is p(x, y, z);

[0220]

[0221] α=ωt;

[0222] d is the distance between the laser radar and point p, β is the pitch angle between the laser radar and point p;

[0223] ω is the angular velocity of the laser radar during scanning;

[0224] The time corresponding to the data of the laser radar collection point p is the tth second, and the relative angle between the radar direction and the initial position at this time is α.

[0225] Specifically, the coarse denoising module is also used to calculate the corresponding local density ρ based on the original 3D point cloud data. i The distance δ from the nearest high density point i ;

[0226] The coarse denoising module is also used to i and δ i Arrange in descending order, based on the preset noise point judgment threshold s, i In the first s% and ρ i The original 3D point cloud data located in the last s% are treated as scattered noise points and removed;

[0227] The coarse denoising module is also used to perform clustering based on the cluster center weights, and remove clusters whose number of points in the clusters after clustering is less than a preset threshold as cluster noise points to obtain coarse denoised point cloud data.

[0228] Specifically, the local density calculation formula of the local density is:

[0229] ρ i =∑ j≠i x(d ij -d c );

[0230] The calculation formula of the distance between the nearest high-density point and the nearest high-density point is:

[0231]

[0232] in,

[0233] d ij For point p j To point p i The distance, d c is the cutoff distance.

[0234] Specifically, the cluster center weight calculation formula of the cluster center weight is:

[0235] γ i =ρ i *δ i .

[0236] Based on the technical solution of the embodiment of the present application, actual devices can be used to replace the above units and modules, that is, the automobile trunk volume measurement system based on laser scanning specifically includes: a laser radar, a rotating table, a rotating table bracket, a control module, a wireless transmission module and a data processing module, and the specific situation is as follows:

[0237] First, the laser radar is installed and fixed on the rotating table. It collects the point cloud data of the inner surface of the car trunk corresponding to the current rotation angle position of the rotating table at a certain sampling interval, and transmits the data to the host computer through the wireless transmission module.

[0238] Second, the rotating table drives the lidar to rotate by pre-setting a suitable rotation speed, thereby realizing the lidar's all-round point cloud data collection of the inner surface of the car trunk.

[0239] Third, the rotating table bracket is used to fix the rotating table and the laser radar installed on the rotating table to keep the reference coordinate system unchanged during the entire data collection process.

[0240] Fourth, the control module controls the operation and stop of the laser radar and the rotating table at the same time to ensure that the radar output data corresponds correctly to the angle of the rotating table.

[0241] Fifth, the wireless transmission module is in full-duplex mode, receiving control commands from the host computer and transmitting them to the control module, while also transmitting the point cloud data collected by the lidar to the host computer.

[0242] Sixth, the data processing module is responsible for coordinate conversion, denoising, simplification and volume calculation of the car trunk point cloud data received by the host computer from the wireless transmission module.

[0243] In summary, the technical solution of the embodiment of the present application can measure the volume of any vehicle model, has strong adaptability and high practical value, and can measure the volume of any confined space, and has good scalability;

[0244] In addition, the measuring device composed of laser radar, rotating table, rotating table bracket, control module and wireless transmission module has a simple structure and low operator requirements, which reduces human operation errors and has high measurement accuracy. It also has a high degree of automation and improves measurement efficiency.

[0245] It should be noted that the specific technical scheme, technical effects and technical problems of the automobile trunk volume measurement system based on laser scanning in the embodiment of the present application are the same as the technical principles of the automobile trunk volume measurement method based on laser scanning in the previous embodiment.

[0246] Referring to Figures 4 to 10 of the specification, a car trunk volume measurement system based on laser scanning includes a two-dimensional laser radar 1, a rotating table 2, a rotating table bracket 3, a control module 4, a wireless transmission module 5, and a host computer 6.

[0247] The detailed settings are as follows:

[0248] The laser radar has two interfaces. The first interface is connected to the control module to control the start and stop of the laser radar. The second interface is connected to the wireless transmission module to transmit the data collected by the laser radar.

[0249] There are four mounting screw holes on the rotating table 2, which are used to fix the laser radar 1 on the rotating table 2 so that the positions of the two are relatively fixed to ensure that the coordinate system does not change.

[0250] The rotating table bracket 3 includes a base and a lifting rod in the middle section. The bracket base is a tripod structure to ensure that the position of the rotating table is fixed, and the lifting rod is used to adjust the height of the rotating table to ensure the accuracy of the measurement result.

[0251] The control module 4 is a single chip microcomputer, which controls the operation of the laser radar 1 and the rotating platform 2 by receiving instructions from the wireless transmission module 5 .

[0252] The wireless transmission module 5 is responsible for sending control instructions to the control module 4 and uploading the collected data to the host computer 6.

[0253] The host computer 6 is responsible for controlling the operation of the entire measuring device and processing point cloud data.

[0254] The acquisition of three-dimensional point cloud coordinate values ​​is as follows Figure 7As shown, the coordinates of point p are related to the distance d from the laser scanning automobile trunk volume measurement system to the laser radar measurement point, the relative pitch angle β between the laser radar and point p, and the relative angle α between the radar orientation and the initial position. The data collected by the laser radar are the distance d and the pitch angle β. The angle α can be calculated from the time and the rotation speed of the turntable.

[0255] refer to Figure 8 , the simplified algorithm steps for denoising point cloud data are shown in the figure:

[0256] Firstly, the improved density peak clustering algorithm is used to roughly denoise the point cloud data to remove the large-scale noise far away from the point cloud body;

[0257] Then, the curvature of each point is calculated, and whether the point is located in a feature-rich area is determined based on the curvature. Different denoising algorithms are used to perform fine denoising on point clouds in different areas to obtain the final denoised point cloud data.

[0258] Furthermore, after data denoising, the data points at the edge of the point cloud are determined and retained without streamlining;

[0259] Then, the data is evenly divided into blocks, the average curvature of each block point is calculated, the feature richness is judged according to the average curvature, and different simplification rates are set for simplification, and finally the denoised and simplified point cloud data is obtained.

[0260] refer to Fig. 9 For the volume calculation of point cloud data, the present embodiment uses a slicing method, which uses a set of equally spaced horizontal planes to cut the point cloud from top to bottom. Assuming there are n horizontal planes with a spacing of h, a series of point cloud slices are obtained, and the slice area S is calculated according to the contour boundary of each point cloud slice. i , then the trunk volume is:

[0261]

[0262] It should be noted that, in this application, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0263] The above is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A method for measuring the volume of a car trunk based on laser scanning, characterized in that: Use the preset LiDAR to scan the interior of the car trunk to obtain original 3D point cloud data; Performing noise reduction and simplification processing on the original three-dimensional point cloud data to obtain simplified point cloud data; Based on the simplified point cloud data, the volume of the car trunk is calculated; The original three-dimensional point cloud data is subjected to noise reduction and simplification processing to obtain simplified point cloud data, including the following steps: Performing large-scale noise removal on the original three-dimensional point cloud data to obtain coarse denoised point cloud data; Calculating curvature based on the coarse denoised point cloud data, and partitioning regions according to the curvature into flat region points and feature region points; Perform fine denoising on flat area points and feature area points respectively to obtain denoised point cloud data; Determine edges based on the denoised point cloud data, and evenly divide the data to obtain a plurality of point cloud blocks; Calculating the average curvature of the point cloud block, and performing point cloud feature partitioning into a flat area and a feature area; Performing a simplification process on the flattened area and the characteristic area to obtain simplified point cloud data; Based on the simplified point cloud data, calculating the volume of the car trunk includes the following steps: A series of point cloud slices are obtained by cutting the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes; Calculate the slice area S according to the contour boundary of each point cloud slice i , and then calculate the trunk volume; The formula for calculating the trunk volume is: Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

2. The method for measuring the volume of a car trunk based on laser scanning according to claim 1, characterized in that: The single-line point cloud data set corresponding to each frame of laser radar data is P = {p1, p2, p3...p n }, n is the number of midpoints of the line; The original three-dimensional point cloud data is p(x, y, z); α=ωt; d is the distance between the laser radar and point p, β is the pitch angle between the laser radar and point p; ω is the angular velocity of the laser radar during scanning; The time corresponding to the data of the laser radar collection point p is the tth second, and the relative angle between the radar direction and the initial position at this time is α.

3. The method for measuring the volume of a car trunk based on laser scanning according to claim 1, characterized in that , performing large-scale noise removal on the plurality of the original three-dimensional point cloud data to obtain coarse denoised point cloud data, comprises the following steps: Based on the original 3D point cloud data, calculate the corresponding local density ρ i The distance δ from the nearest high density point i ; P i and δ i Arrange in descending order, based on the preset noise point judgment threshold s, i In the first s% and ρ i The original 3D point cloud data located in the last s% are treated as scattered noise points and removed; Clustering is performed based on the cluster center weights, and clusters whose number of points in the clusters after clustering is less than the preset threshold are removed as cluster noise points to obtain coarse denoised point cloud data.

4. The method for measuring the volume of a car trunk based on laser scanning according to claim 3, characterized in that: The local density calculation formula of the local density is: r i =∑ j≠i x(d ij -d c ); The calculation formula of the distance between the nearest high-density point and the nearest high-density point is: in, d ij For point p j To point p i The distance, d c is the cutoff distance.

5. The method for measuring the volume of a car trunk based on laser scanning according to claim 4, characterized in that: The cluster center weight calculation formula of the cluster center weight is: c i =ρ i *d i 。 6. A car trunk volume measurement system based on laser scanning, characterized in that: The system comprises: LiDAR, which is used to scan the interior of the car trunk to obtain raw 3D point cloud data; A data reduction element is used to reduce noise and reduce the original three-dimensional point cloud data to obtain reduced point cloud data; A volume calculation unit, which is used to calculate the volume of the car trunk based on the simplified point cloud data; The data reduction elements include: A coarse denoising module, which is used to remove large-scale noise from the original three-dimensional point cloud data to obtain coarse denoised point cloud data; A coarse denoising area point division module, which is used to calculate the curvature based on the coarse denoising point cloud data, and divide the area into flat area points and feature area points according to the curvature; A fine denoising module is used to perform fine denoising on flat area points and feature area points respectively to obtain denoised point cloud data; A point cloud block division module, which is used to determine the edge based on the denoised point cloud data and divide it evenly to obtain multiple point cloud blocks; A fine denoising area division module, which is used to calculate the average curvature of the point cloud block and perform point cloud feature partitioning into a flat area and a feature area; A fine denoising and streamlining module, which is used to perform streamlining processing on the flattened area and the feature area to obtain streamlined point cloud data; The volume calculation unit comprises: A point cloud slicing module, which is used to cut the point cloud corresponding to the simplified point cloud data from top to bottom using a set of equally spaced horizontal planes to obtain a series of point cloud slices; Volume calculation module, which is used to calculate the slice area S according to the contour boundary of each point cloud slice i , and then calculate the trunk volume; The formula for calculating the trunk volume is: Among them, V is the volume of the trunk, h is the distance between the horizontal planes, and n is the number of point cloud slices.

7. The automobile trunk volume measurement system based on laser scanning as claimed in claim 6, characterized in that: The system further comprises: A rotating platform is used to control the laser radar to rotate at a preset angular velocity.

Images