作者: Jae-Seol Lee , Jun-Hyeong Jo , Tae-Hyoung Park
DOI: 10.1109/TITS.2019.2903529
关键词:
摘要: An effective method to segment vehicles and roads is proposed for autonomous using low-channel 3D lidar. The distance-view transformation newly overcome the low density of top-view data In addition, a dilated convolution structure expand receptive field convolutional neural network. network improves accuracy segmentation. experimental results are presented verify usefulness method.