深度学习中的单阶段车辆检测算法综述

随着深度学习的发展,基于深度学习的车辆检测算法性能不断被提升,在构建智能交通体系方面发挥重要作用.单阶段目标检测模型因其检测速度的优越性,被广泛应用于车辆实时检测.为了综合分析基于深度学习的单阶段车辆检测算法相关改进及应用,分别对比了各类常用单阶段车辆检测算法,列举其改进措施以及在车辆检测方面存在的问题;重点阐述了基于...

Review of One-Stage Vehicle Detection Algorithms Based on Deep Learning

With the development of deep learning, the performance of vehicle detection algorithm based on deep learning is continuously improved and plays an important role in building intelligent transportation system. One-stage target detection model is widely used in real-time vehicle detection in video stream due to its advantages in detection speed. In order to analyze the relevant improvements and applications of one-stage vehicle detection algorithm based on deep learning comprehensively, various common one-stage vehicle detection algorithms are compared by listing their improvements and problems left to be solved in vehicle detection, and then emphatically describes the relevant improvements and application fields of vehicle algorithm based on common one-stage framework. Finally, the data sets of vehicle detection are introduced briefly, the problems and difficulties to be solved in vehicle detection at the present stage are analyzed, and the feature research direction is put forward.