Key points are not available for this paper at this time.
As autonomous driving has become more popular, research on the technology has proliferated. There are many schemes to realize autonomous driving, including visual perception, lidar, and radar. Among them, visual perception implements advanced neural networks and transfer learning to achieve excellent performance. This article mainly discusses the advantages of visual perception, the challenges it faces, and possible solutions, while comparing it with other schemes from various aspects at the same time. The approaches to the study mainly involve literature review and comparative analysis. The article concludes that visual perception will thrive and dominate the autonomous driving industry in the future. The prospect of development is also investigated, and some potential solutions that can improve the performance of autonomous driving are provided in the passage. This comprehensive analysis underscores the pivotal role of the visual perception in advancing autonomous driving technology, highlighting its cost-effectiveness, scalability, and future potential in transforming transportation.
Haotang Wang (Mon,) studied this question.