Lightweight point cloud target detection method oriented to representative point self-attention mechanism
A technology of target detection and representative point, which is applied in the field of target detection based on point cloud data, can solve the problems of lack of local feature information of representative points, insufficient feature expression ability, occupying more computing resources, etc., so as to improve the detection accuracy and realize the The effect of reducing weight and reducing the waste of computing resources
Pending Publication Date: 2022-01-28
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] These methods have the following disadvantages: (1) There is a problem of repeated calculation at some points, and the amount of model parameters is increased at the same time, and it takes longer time and takes more computing resources for training
(2) In the process of selecting representative points, the characteristic connection between representative points and surrounding adjacent points may be ignored, and there is a problem of missing local feature information of representative points
(3) Due to the sparsity of point cloud data, when the point cloud data is in the process of forward propagation, it will become more and more sparse, resulting in insufficient feature expression ability
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0041] The flow of the lightweight point cloud object detection method for representative point self-attention mechanism in this embodiment is as follows: figure 2 As shown, it includes the following steps:
[0042] S1. Read point cloud data;
[0043] In this step, the point cloud data in a single scene is composed of an unordered point cloud collection, {x 1 , x 2 ,...,x n}, the information of each point is composed of three-dimensional coordinates and reflectivity. Since the number of point clouds contained in each scene is different, in order to ensure the consistency of input data in multiple scenes, therefore, when reading point cloud data Consistent sampling of the number of point clouds is required.
[0044] S2. Perform FPS sampling on the point cloud data to obtain representative points; in this step, use the farthest point sampling (Farthest Point Sampling, FPS) on the read point cloud data to obtain 8192 (the best value obtained in experiments) The point cloud, a...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a target detection technology based on point cloud data, and discloses a lightweight point cloud target detection method oriented to a representative point self-attention mechanism. The method reduces the parameter quantity of a model and improves the training convergence rate and the detection accuracy of the model. The method comprises the following steps: S1, reading point cloud data; S2, performing FPS sampling on the point cloud data to obtain representative points; S3, performing multi-layer K neighbor feature extraction on the representative points to obtain feature vectors of the representative points; and S4, performing classification detection according to the feature vector of each representative point to obtain the probability that the point contains the target.
Description
technical field [0001] The invention relates to a target detection technology based on point cloud data, in particular to a lightweight point cloud target detection method oriented to a representative point self-attention mechanism. Background technique [0002] 3D computer vision technology is changing with each passing day and developing rapidly. As an important branch of computer vision, target detection has broad application prospects in social life and industry. It is a popular field in both academia and industry. Because point cloud data is not easily affected by factors such as light and dust, while traditional images may suffer from visual interference factors such as weather, point cloud data has been widely used in target detection tasks in 3D scenes. By using point Cloud data obtains three-dimensional space information, improves the effect of three-dimensional object detection, and has a wide range of application prospects and application value. [0003] In the ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/2415G06F18/253Y02D10/00
Inventor 朱大勇罗光春赵太银陈爱国潘海涛曹申健
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA