基于强化语义流场和多级特征融合的道路场景分割方法  

Enhance Semantic Flow Field and Multilevel Feature Fusion Network for Road Scene Segmentation

作　　者：项建弘[1,2] 刘茁 王霖郁 钟瑜[3] XIANG Jianhong;LIU Zhuo;WANG Linyu;ZHONG Yu(College of Information&Communication Engineering,Harbin Engineering University,Harbin 150001,China.;Key Laboratory of Advanced Ship Communication and Information Technology,Harbin Engineering University,Harbin 150001,China;Southwest Electronic Technology Research Institute of China,Chengdu 610036,China)

机构地区：[1]哈尔滨工程大学信息与通信工程学院,哈尔滨150001 [2]哈尔滨工程大学先进船舶通信与信息技术重点实验室,哈尔滨150001 [3]中国西南电子技术研究所,成都610036

出　　处：《数据采集与处理》2022年第2期426-436,共11页Journal of Data Acquisition and Processing

摘　　要：自动驾驶是目前计算机视觉任务中难度较大的一类任务,而道路场景下的语义分割是自动驾驶的核心技术之一。本文针对经典分割网络中分辨率恢复方式简单,导致细节信息不完整、目标边缘模糊的问题,提出一种基于强化语义流场的上采样方法。该方法通过学习相邻特征图之间的语义流场,使生成图语义信息更细致,边界处更清晰。同时针对道路场景中目标尺度变化处理困难、小目标难以识别的问题,提出一种新的多级特征融合方法,充分融合深层语义信息与浅层细节信息,以适应不同尺度的目标。本文采用CamVid为数据集进行实验,并进行数据增强。实验表明本文提出的两种方法均显著提升了准确度,整体网络与PSPNet、Deeplabv3+等多种模型相比,准确率更高,分割效果更接近真实值。Automatic driving is one of the most difficult tasks in computer vision,and semantic segmentation in road scenes is one of the core technologies of automatic driving.This paper proposes an upsampling method based on enhanced semantic flow field,which can make the semantic information of the generated graph more detailed and the boundary clearer by learning the semantic flow field between adjacent feature graphs.At the same time,aiming at the difficulty of processing target scale changes and identifying small targets in road scenes,a new multi-level feature fusion method is proposed,which fully integrates deep semantic information and shallow detail information to adapt to targets of different scales.In this paper,CamVid is taken as the data set and data enhancement is carried out.Experiments show that both methods proposed in this paper bring significant improvement in accuracy.Compared with PSPNet,Deeplabv3+and other models,the overall network has higher accuracy and the segmentation effect is closer to the real value.

关 键 词：深度学习 语义分割 道路场景 强化语义流场 多级特征融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]