An indoor mapping method based on slam single-line radar and inertial measurement unit

Abstract

The application discloses an indoor mapping method based on a SLAM single-line radar and an inertial measurement unit, and comprises the following steps: step 1, acquiring point cloud data sets measured by the single-line radar and Euler angle pose data corresponding to each point cloud data and measured by the inertial measurement unit; step 2, correcting the pose of the point cloud data sets based on the Euler angle obtained in step 1 to obtain point cloud data after pose correction; step 3, constructing a preliminary indoor building plane point cloud map by using the point cloud data after pose correction; and step 4, fitting a two-dimensional line frame based on a least square method segmented linear function, fitting a segmented straight line for the point cloud data after pose correction in step 2, and finally obtaining a precise indoor building plane line frame standard layout map. The application solves the problem that different coordinate systems lead to complex data processing, and overcomes the problem of point cloud data distortion of the single-line radar in a complex environment by detecting the Euler angle of the single-line radar through the inertial measurement unit.

Description

Technical Field

[0001] This invention relates to the field of indoor mapping, and more particularly to an indoor mapping method based on SLAM single-line radar and inertial measurement unit. Background Technology

[0002] Digital models of architectural spaces serve as a universal foundation for renovation, inspection, and decoration. In the digital age, three-dimensional spatial models have become a common and crucial basis for building renovation design, safety verification calculations, decoration plans, and intelligent visual operation and maintenance. They are also an important component in expanding urban digital spaces. On-site reverse modeling is an effective means of obtaining a realistic building model. Due to differences in construction age, technology, and quality, buildings vary greatly in form and internal structure. Furthermore, with the passage of time and data transfer, original drawings are often missing or do not match reality, making it difficult to obtain a realistic building model.

[0003] Currently, in fields such as inspection and identification, and decoration and renovation, modeling is mostly done through manual distance measurement and foot measurement. However, this method generally suffers from problems such as high roughness, low efficiency, and limited application scenarios. The development of emerging technologies such as laser scanning has greatly improved the accuracy and application scope of reverse modeling of buildings, and it has already been applied in fields such as the protection of ancient buildings. However, the application of existing laser scanning modeling technology still faces significant limitations. First, the barrier to entry is high; the massive point cloud data obtained from panoramic scanning places extremely high demands on equipment performance and backend computing power, severely restricting its widespread application. Second, efficiency is low; the entire process from scanning the point cloud on-site to obtaining the model through subsequent calculations is time-consuming, failing to demonstrate any time advantage over traditional manual distance measurement and drawing. Third, the equipment is expensive, making it unsuitable for ordinary engineering applications. Summary of the Invention

[0004] The main objective of this invention is to overcome the shortcomings of the above-mentioned technologies and provide an indoor mapping method based on SLAM single-line radar and inertial measurement unit, which can significantly reduce the time required for on-site inspection and mapping of existing buildings, reduce the amount of calculation, and reduce labor and equipment costs.

[0005] To achieve the above objectives, this invention provides an indoor mapping method based on SLAM single-line radar and an inertial measurement unit, comprising the following steps:

[0006] Step 1: Obtain the point cloud dataset measured by single-line radar, and the Euler angle pose data measured by the inertial measurement unit corresponding to each point cloud dataset;

[0007] Step 2: Based on the Euler angles obtained in Step 1, perform pose correction on the point cloud dataset to obtain the pose-corrected point cloud data;

[0008] Step 3: Use the pose-corrected point cloud data to construct a preliminary building interior plan point cloud map;

[0009] Step 4: Fit a two-dimensional wireframe using a piecewise linear function based on the least squares method. Perform piecewise linear fitting on the point cloud data after pose correction in Step 2 to finally obtain an accurate standard layout drawing of the building's interior plan wireframe.

[0010] Preferably, in step 1, the lidar and the inertial measurement unit are initially calibrated on the same Z-axis, and the point cloud dataset and the Euler angle pose timestamp are matched point-to-point according to the most recent time.

[0011] Preferably, step 2 is achieved by correcting the coordinates using a unit real quaternion, and the specific implementation method is as follows:

[0012] Using the scanned data point A = (x, y, z) as the rotation base point, and given the Euler angles of rotation as: yaw angle γ for the Z-axis, roll angle α for the X-axis, and pitch angle β for the Y-axis, assume the unit real quaternion corresponding to the 3D rotation transformation is... The corrected point after rotation is A′=(x′,y′,z′), and the relationship between Q and Euler angles is:

[0013]

[0014] The relationship between A and A′ is:

[0015]

[0016] A′ can be calculated using the above formulas (1) and (2). The calculation starts from the first point cloud data point obtained by the lidar and continues until the pose correction of all subsequent point cloud data points is completed.

[0017] Preferably, step 3 is achieved by retaining the x and y values ​​of the corrected point cloud dataset from step 2.

[0018] Preferably, in step 4, the specific implementation method for piecewise linear fitting is as follows:

[0019] Given two-dimensional data points (x1, y1), ..., (x i ,y i Let the empirical function fitted to the first straight line be y = ax + b, and denote the sum of squared residuals. Considering the minimum value of the sum of squared residuals, we can find a and b by taking the partial derivatives with respect to a and b respectively:

[0020]

[0021] Take two-dimensional data points (x) i ,y i ), (x i+1 ,y i+1 ), (xi+2 ,y i+2 ), correspondingly find the empirical function y = a′x + b′ for the second straight line, let a = tanφ and a′ = tanψ, the angle between the two lines is Then we have:

[0022]

[0023] In equation (4), if If the angle is less than 15°, then (x) i+1 ,y i+1 If the points are placed in front, the linear equation is refitted; If the angle is greater than 15°, the equation of the first fitted line is y = ax + b. Then, y = a′x + b′ is taken as the new first line equation, and (x i+1 ,y i+1 ), (x i+2 ,y i+2 ), (x i+3 ,y i+3 Repeat the above steps until all point cloud data is fitted, and you will get all the fitting results of the piecewise straight lines, and finally obtain an accurate architectural interior plan layout.

[0024] Due to the adoption of the above technical solution, the beneficial effects of the present invention are as follows:

[0025] (1) The indoor drawing method based on SLAM single-line radar and inertial measurement unit provided by the present invention can significantly reduce the labor cost and workload of existing building indoor drawing, improve measurement accuracy, reduce equipment cost, and obtain the building indoor plan layout drawing by using the pose correction of unit real quaternions and the least squares piecewise straight line fitting method.

[0026] (2) By initially calibrating the inertial measurement unit and the SLAM single-line lidar on the same Z-axis, the problem of complex data processing caused by the different coordinate systems of the two is solved. At the same time, by detecting the Euler angle of the single-line lidar through the inertial measurement unit, the problem of point cloud data distortion of the single-line lidar in complex environments is overcome. Attached Figure Description

[0027] Figure 1 This is a flowchart of the indoor mapping method based on SLAM single-line radar and inertial measurement unit according to the present invention;

[0028] Figure 2 This is a comparison image of point cloud data obtained after pose correction in an embodiment of the present invention.

[0029] Figure 3 This is a schematic diagram of the final architectural interior plan line frame obtained in an embodiment of the present invention. Detailed Implementation

[0030] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments:

[0031] Combination Figure 1 As shown, an indoor mapping method based on SLAM single-line radar and inertial measurement unit includes the following steps:

[0032] Step 1: Obtain the point cloud dataset measured by the single-line radar, as well as the Euler angle pose data measured by the inertial measurement unit corresponding to each point cloud dataset. Perform initial calibration on the lidar and the inertial measurement unit so that they are on the same Z-axis. Match the point cloud dataset and the Euler angle pose timestamps according to the most recent time.

[0033] Step 2: Based on the Euler angles obtained in Step 1, the pose of the point cloud dataset is corrected by substituting each point cloud data and its corresponding Euler angle data into formula (1) and formula (2) to obtain the corrected point cloud dataset.

[0034]

[0035]

[0036] Step 3: Using the pose-corrected point cloud data, construct a preliminary building interior plan point cloud map. Retain the x and y values ​​corresponding to the corrected point cloud dataset from Step 2, recalculate the corresponding l′ and θ′ values ​​according to x′=l′cosθ′ and y′=l′sinθ′, and display them, such as... Figure 2 As shown.

[0037] Step 4: Fit the point cloud data after pose correction in Step 2 using the least squares piecewise linear fitting method to obtain a two-dimensional wireframe. The specific calculation is as follows:

[0038] In this embodiment, the first three laser data points are selected as the data points for the first straight line fitting. The equation of the straight line y = ax + b can be calculated using formula (3).

[0039]

[0040] Furthermore, select two-dimensional data points (x3, y3), (x4, y4), and (x5, y5), and repeat formula (3) to find the empirical function y = a′x + b′ for the second straight line. Let a = tanφ and a′ = tanψ, and the angle between the two lines is... Then we have:

[0041]

[0042] In equation (4), if If the angle is less than 15°, then place (x4, y4) among the previous points and refit the straight line equation; if For angles greater than 15°, the equation of the first fitted line is y = ax + b. Then, y = a′x + b′ is used as the new first line equation. The above steps are repeated for (x4, y4), (x5, y5), and (x6, y6) until all point cloud data is fitted. This yields the complete fitting results for the piecewise lines, ultimately resulting in an accurate architectural interior plan layout. Figure 3 As shown.

[0043] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made under the concept of the present invention using the description and drawings of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. An indoor mapping method based on SLAM single-line radar and inertial measurement unit, characterized in that, Includes the following steps: Step 1: Obtain the point cloud dataset measured by single-line radar, and the Euler angle pose data measured by the inertial measurement unit corresponding to each point cloud dataset; Step 2: Based on the Euler angles obtained in Step 1, perform pose correction on the point cloud dataset to obtain the pose-corrected point cloud data; Step 3: Use the pose-corrected point cloud data to construct a preliminary building interior plan point cloud map; Step 4: Fit a two-dimensional wireframe using a piecewise linear function based on the least squares method. Fit a piecewise linear function to the point cloud data after pose correction in Step 2 to obtain an accurate building interior floor plan. In step 4, the specific implementation method for piecewise line fitting is as follows: Given two-dimensional data points ... Let the empirical function for fitting the first straight line be... Record the sum of squared residuals Considering the minimum value of the sum of squared residuals, respectively for and By taking the partial derivative, we can find the answer. , : 、 （3） Take two-dimensional data points , , The empirical function for the second straight line is then obtained. ,make , The angle between the two lines is Then we have: （4） In equation (4), if If less than 15°, then... Refit the linear equation using the points mentioned earlier; if For angles greater than 15°, the equation of the fitted straight line for the first segment is: , and then As the first new straight line equation, , , Repeat the above steps until all point cloud data is fitted, obtain all fitting results for the piecewise straight lines, and finally obtain an accurate standard layout drawing of the building's interior plan wireframe.

2. The indoor mapping method based on SLAM single-line radar and inertial measurement unit according to claim 1, characterized in that: In step 1, the lidar and inertial measurement unit are initially calibrated on the same Z-axis, and the point cloud dataset and Euler angle pose timestamp are matched point-to-point according to the most recent time.

3. The indoor mapping method based on SLAM single-line radar and inertial measurement unit according to claim 1, characterized in that: Step 2 is achieved by correcting the coordinates using a unit real quaternion, and the specific implementation method is as follows: To scan data points As the base point of rotation, the Euler angles of rotation are known to be: For the yaw angle along the Z-axis For the roll angle on the X-axis Pitch angle for the Y-axis ; Assume the unit real quaternion corresponding to the three-dimensional rotation transformation is The correction point after rotation is The relationship between Q and Euler angles is: （1） but and The relationship between them is: （2） The above formula can be used to calculate... The calculation starts from the first point cloud data point obtained by the lidar and continues until the pose correction of all subsequent point cloud data points is completed.

4. The indoor mapping method based on SLAM single-line radar and inertial measurement unit according to claim 3, characterized in that: Step 3 is achieved by retaining the x and y values ​​of the corrected point cloud dataset from Step 2.

Images