基于多模态融合的自动驾驶环境感知算法优化研究

打开文本图片集

中图分类号：U471 收稿日期：2025-12-13

DOI： 10.19999/j.cnki.1004-0226.2026.02.013

Research on Optimization of Autonomous Driving Environmental Perception Algorithm Based on Multimodal Fusion

Bai Huizhen Dou Feijuan Li Yinxia Xi'anAutomobile Vocational University，Xi'an 7106oo，China

Abstract：Environmentalperceptionisacorecomponentofautonomousdrivingsystems，andmultimodalfusiontechnologypro videsanewpathforimprovingtheacuracyandperformanceofperception.Thispaperfocusesontheresearchofperceptionalgorithms inautonomousdrivingscenarios.Basedonthefusionofmulti-sourcedatasuchasimagsandmillimeterwaveradarsanefcientfusion modelisconstructedandthefeatureextractionanddecisionmakingmechanismsareoptimized.Theresearchresultsshowthatthepro posedmethodhasigherecognitioaccuracyandprocessngabilityinariouscomplexeviroments，rovidingteoreticalsupport and technical reference for the practical application of environmental perception systems.

Key words：Autonomous driving;Multimodal fusion;Environmental perception;Algorithm optimization

1前言

自动驾驶技术的安全性和可靠性高度依赖于环境感知系统的性能。（剩余5260字）