A method and system for underwater terrain matching and positioning of three-dimensional imaging sonar point cloud

Through the underwater terrain matching positioning method of three-dimensional imaging of sonar point clouds, deep learning technology is used to extract the terrain key points of sonar point clouds, solving the problems of low accuracy and insufficient robustness of traditional underwater terrain matching positioning algorithms, and realizing the High-precision positioning under arbitrary transformation.

CN114706087BActive Publication Date: 2024-11-05NAT UNIV OF DEFENSE TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210352439.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-04
Publication Date
2024-11-05
Estimated Expiration
2042-04-04

AI Technical Summary

Technical Problem

The traditional underwater terrain matching positioning algorithm has problems such as low positioning accuracy, low robustness, and can only deal with rigid transformations, and has limited application scope.

Method used

The underwater terrain matching positioning method of three-dimensional imaging sonar point clouds is adopted, and the terrain key points of sonar point clouds are extracted through key point extraction networks and deep learning technology, and the matching cost between real-time point cloud frames and maps is calculated, so as to predict the position deviation. Movement to achieve high-precision positioning under any transformation.

Benefits of technology

It improves the accuracy and robustness of underwater terrain matching positioning, can handle non-rigid transformations, expands the scope of application, reduces the dependence of manually set feature extraction modes, and improves the real-time and reliability of positioning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114706087B_ABST
    Figure CN114706087B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and system for underwater terrain matching and positioning of three-dimensional imaging sonar point clouds, the method comprising: S1, sampling the sonar point cloud frames collected online, increasing the number of points to N, and cutting the pre-stored large-scale sonar point cloud map into a point cloud X with N points; S2, transforming L randomly generated transformation matrices {T 1 ,…,T L},T L ∈SE(3) acts on the point cloud X = [X 0 ,…,X N ]∈R 3×N , obtain L pairs of data #imgabs0# to train the key point extraction network; S3, select multiple neighborhood points of each key point, and extract features from the neighborhood points, respectively extract features from the pre-stored sonar point cloud topographic map and the online collected sonar point cloud topographic map with the neighborhood range of each key point; S4, calculate the matching cost between the real-time point cloud frame and the map according to the key points; S5, calculate the position offset of the sonar point cloud frame and the map under the prediction in real time according to the matching cost; S6, define the loss according to the square of the distance between the predicted position offset and the ground truth value. The present invention can improve the matching effect, reduce errors and accurately locate.
Need to check novelty before this filing date? Find Prior Art