Patent · US Active

Method and device for mapping three-dimensional (3D) point cloud model based on deep learning

US12175691B1 · kind B1 · utility

1Cited by
4References
7Claims
0Family size

Assignees

Inventors

Key dates

Filing dateDec 28, 2023
Grant dateDec 24, 2024
Priority date
Expiry dateDec 28, 2043

Classification

  • Technology area (CPC Y)Emerging Cross-Sectional Technologies
  • CPC primaryY02A90/10
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A method for mapping a three-dimensional (3D) point cloud model based on deep learning includes: extracting features of image data by using a convolutional neural network; extracting features of point cloud data by using a PointNet point cloud processing network, and constructing a 3D model of the point cloud data by using a triangular mesh; aligning the 3D model and an image spatially and temporally; performing projection mapping on the aligned 3D model and image, and superimposing location information in the point cloud data onto texture information of the image to obtain a first fused image; fusing the features of the point cloud data and the features of the image data to obtain a second fused image with a fused feature; and superimposing the first fused image onto the second fused image, and obtaining a superimposition weight by using the convolutional neural network, to generate a mapped 3D model.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.