A scene reconstruction method and device, vehicle and storage medium

By sorting and moving overlapping 3D models according to the distance between objects and vehicles during vehicle scene reconstruction, the misjudgment problem caused by object overlap in existing technologies is solved, thereby improving the accuracy and safety of scene reconstruction.

CN122415892APending Publication Date: 2026-07-17CHONGQING CHANGAN AUTOMOBILE CO LTD

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHONGQING CHANGAN AUTOMOBILE CO LTD
Filing Date
2026-05-27
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

In existing in-vehicle scene reconstruction technologies, the size ratio difference between the pre-built 3D model and the actual object causes multiple object 3D models in the scene model to overlap, which cannot accurately reflect the real positional relationship between objects outside the vehicle. This leads to the driver misjudging the relative position and distance of objects, posing a serious driving safety hazard.

Method used

By acquiring the initial scene model of the vehicle in the current frame, the distance between each object and the vehicle is determined, and the objects are traversed in order of distance from near to far. Overlapping 3D models are moved to eliminate overlap, ensuring the accuracy and safety of the reconstructed scene model.

Benefits of technology

It achieves a high degree of matching between the reconstructed scene and the actual vehicle external scene, eliminating safety hazards caused by driver misjudgment due to overlapping models, and improving the accuracy of in-vehicle scene reconstruction and driving safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122415892A_ABST
    Figure CN122415892A_ABST
Patent Text Reader

Abstract

本申请涉及一种场景重构方法、装置、车辆及存储介质,方法包括:获取当前帧车辆对应的初始场景模型;基于初始场景模型中各个三维模型,确定各个对象与车辆之间的距离,及基于所得到的各个距离,确定各个对象之间的处理顺序,距离越小的对象,在处理顺序中越靠前;基于处理顺序,按照如下方式依次遍历各个对象:在当前遍历对象的三维模型与任一目标三维模型存在重叠时,将重叠的目标三维模型进行移动,以消除重叠,目标三维模型为未遍历对象的三维模型;在遍历完所有对象后,将消除重叠后的初始场景模型确定为车辆对应的重构场景模型,以完成车辆的场景重构。本申请提高了车载场景重构的准确性与驾驶安全性。
Need to check novelty before this filing date? Find Prior Art