Semantic aerial view visual relocation method and device in non-exposed scene, electronic equipment, storage medium and program product

By using a semantic-geometric fusion mechanism and an edge-cloud collaborative architecture, a dense 3D semantic point cloud is generated and semantic bird's-eye view relocalization is performed. This solves the problems of positioning accuracy and stability in non-exposed scenarios, achieving high-precision and robust visual relocalization, which is suitable for intelligent transportation and underground inspection.

CN121505029APending Publication Date: 2026-02-10BEIJING UNIV OF POSTS & TELECOMM
0 Cites 5 Cited by

Patent Information

Application Number
CN202511595280.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-03
Publication Date
2026-02-10

AI Technical Summary

Technical Problem

In non-exposed scenarios where GNSS signals are limited or obstructed, traditional visual relocation methods suffer from insufficient positioning accuracy, poor semantic consistency, low utilization of geometric constraints, and limited real-time performance, making it difficult to achieve high-precision and stable positioning in complex environments.

Method used

By constructing a semantic-geometric fusion mechanism, a dense 3D semantic point cloud is generated using a pre-trained deep semantic detection model and a voxel differentiable rendering model. Combined with the short-term trajectory information of the inertial measurement unit, joint mapping and relocalization of semantic and geometric information are achieved. A semantic mask-guided reciprocal matching strategy is adopted to improve matching stability, and real-time deployment is achieved through an edge-cloud collaborative architecture.

Benefits of technology

It achieves centimeter-level positioning accuracy in non-exposed environments, suppresses matching drift under conditions of varying illumination and complex structures, possesses sub-second real-time performance and high robustness, outputs interpretable semantic bird's-eye view maps, and supports applications such as intelligent transportation and underground inspection.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a semantic aerial view visual relocation method and device in a non-exposed scene, electronic equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring a multi-view image sequence under a non-exposed scene (such as a tunnel, an underground pipe gallery or an underground parking lot); semantic recognition is carried out based on a pre-trained semantic target detection model, and spatial consistency semantic features are extracted through a semantic-geometric dual-channel fusion mechanism combining a semantic mask and geometric constraints; the method comprises the following steps of: realizing three-dimensional reconstruction by using a voxel micro-renderable modeling method (VGGT), and generating a dense three-dimensional semantic point cloud fusing semantics and a geometric structure; two-dimensional semantics are mapped to a three-dimensional space through a projection and back projection relation, and point cloud semantics are endowed; main structure planes such as the ground, the left wall surface and the right wall surface are extracted, and a two-dimensional semantic aerial view with semantic annotation is generated; and pose estimation is carried out based on a reciprocal matching strategy guided by a semantic mask, so that visual repositioning with high precision, high robustness and semantic interpretability is realized. The method breaks through the problems of low precision, sparse features and poor semantic consistency of traditional visual repositioning in a non-exposed environment, and can be widely applied to the fields of intelligent transportation, underground inspection and unmanned system positioning.
Need to check novelty before this filing date? Find Prior Art