Laser radar point cloud registration method for urban dynamic environment

A laser radar point cloud registration method for an urban dynamic environment, comprising the following steps:

• S1:
• establishing a data set of a point cloud registration network;
• S2:
• constructing the point cloud registration network based on multi-task learning:
• the point cloud registration network based on the multi-task learning comprises a static reference target feature extraction module and a point cloud segmentation and registration module, wherein
• 1) the static reference target feature extraction module
• a 3D point cloud is converted into a 2D distance image by means of spherical projection, wherein point cloud coordinates (x, y, z) are converted into image coordinates (u, v):
• (
• u
• v
• )
• =
• (
• 1
• 2
• [
• 1
• -
• arctan
• ⁡
• (
• y
• ,
• x
• )
• ⁢
• π
• -
• 1
• ]
• ⁢
• w
• [
• 1
• -
• (
• arcsin
• (
• z
• ,
• r
• -
• 1
• )
• +
• f
• up
• )
• ⁢
• f
• -
• 1
• ]
• ⁢
• h
• )
• in the formula, w and h represent a width and height of a distance image, r represents a distance of each point, r=√{square root over (x2=y2+z2)}, f represents a vertical field of view of a laser radar, and fup represents an upward angle of f;
• the coordinates (x, y, z), distance r, and reflectivity i of each frame of point cloud are stored in a distance image with a size of [w×h×5];
• the static reference target feature extraction module inputs distance images at moments k−1 and k, and outputs static reference target features;
• the two distance images are decomposed into n image patches, n=(2×w×h)/({dot over (w)}×{dot over (h)}), a size of each image patch is [{dot over (w)}×{dot over (h)}×5], each image patch is expanded into a J-dimensional vector, J=5×{dot over (w)}×{dot over (h)}, and the two distance images are represented as a sequence {dot over (Q)}=[q1, q2, . . . , qn], q∈
• static reference target features are extracted by using a Transformer encoder, a vector dimension processed by the Transformer encoder is K, the sequence {dot over (Q)} is converted into a sequence {umlaut over (Q)}=[q1E, q2E, . . . , qnE] through a learnable linear projection E∈, and a learnable registration vector q0∈ is added to the sequence {umlaut over (Q)}, to obtain a sequence {umlaut over ({dot over (Q)})}=[q0, q1E, q2E, . . . , qnE];
• a learnable position vector P∈ is added to the sequence {dot over ({umlaut over (Q)})} to obtain a Transformer encoder input Q0=[q0, q1E, q2E, . . . , qnE]+P;
• the Transformer encoder has L modules, each of the modules includes a multi-head self-attention (MHSA) layer and a multi-layer perceptron (MLP) layer, layer normalization (LN) is performed before data enters each layer, residual connections are used to fuse an output and input of each layer, and a feature calculation process of a module 1 is as follows:
• Q
• 1
• ′
• =
• MSA
• ⁡
• (
• LN
• (
• Q
• 1
• -
• 1
• )
• )
• +
• Q
• 1
• -
• 1
• ,
• 1
• =
• 1
• ,
• 2
• ,
• ...
• ,
• L
• Q
• 1
• =
• MLP
• ⁡
• (
• LN
• (
• Q
• 1
• ′
• )
• )
• +
• Q
• 1
• ′
• ,
• 1
• =
• 1
• ,
• 2
• ,
• ...
• ,
• L
• a static reference target feature QL is obtained based on Q0 through the Transformer encoder;
• 2) the point cloud segmentation and registration module
• the static reference target feature QL is a sequence of n+1 K-dimensional vectors, a vector with an index 0 is used to estimate a transformation matrix Zk-1, k=[Rk-1, ktk-1, k] at the moments k−1 to k, vectors with indices 1 to 0.5n are used to obtain a static reference target segmentation result at the moment k−1, vectors with indices 0.5n+1 to n are used to obtain a static reference target segmentation result at the moment k, and the same operation, i.e., parameter sharing, is performed for segmentation at both the two moments;
• the transformation matrix Zk-1, k is obtained through a designed neural network, the neural network comprises one input layer, two hidden layers, and one output layer, and a Tanh activation function is added to neurons in the hidden layers;
• a process of obtaining a static reference object segmentation result comprises three steps:
• first converting 0.5n K-dimensional vectors into 0.5n J-dimensional vectors through linear projection, then transforming and sorting the 0.5n J-dimensional vectors to obtain a feature map with a size of [w×h×5], and finally processing the feature map using a 1×1 convolution kernel and a Softmax function to obtain the static reference object segmentation results at the moments k−1 and k;
• S3:
• training the designed point cloud registration network;
• and
• S4:
• using the trained network to estimate a transformation matrix of two frames of point clouds.