Abstract：Semantic segmentation of indoor scenes has always been an important direction in the field of deep learning semantic segmentation. The main problems of indoor scene segmentation are many semantic categories, many object classes will block each other, and some classes have high similarity. Aiming at these problems, Proposed a method for semantic segmentation of indoor scenes which is based on the BiSeNet (bilateral segmentation network), this method introduces a hollow pyramid pooling layer and a multi-scale feature fusion module. The features are fused to obtain enhanced content features, which improves the models performance for semantic segmentation of indoor scenes. The MIoU performance of this method on the indoor scene dataset in ADE20K increased by 23.5% compared toSegNet and 3.5% compared to before model.
［1］Kostavelis I, Gasteratos A. Semantic mapping for mobile robotics tasks: A survey［J］. Robotics and Autonomous Systems, 2015, 66(4): 86-103.
［2］Pronobis A, Jensfelt P. Large-scale semantic mapping and reasoning withheterogeneous modalities［C］//2012 IEEE International Conference on Robotics and Automationin, IEEE, SaintPaul, MN, USA, 3515-3522.
［3］胡最, 闫浩文. 基于语义模型的地图符号库设计研究［J］. 地理空间信息, 2008, 6(2): 100-102.
Hu Z, Yan H W. Research on Map Symbol Library Design Based on Semantic Model ［J］. Geospatial Information, 2008, 6(2): 100-102.
［4］Schnberger J L, Marc P, Geiger A, et al. Semantic Visual Localization［C］//The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, 6896-6906.
［5］吴凡, 闵华松. 一种实时的三维语义地图生成方法［J］. 计算机工程与应用, 2017, 53(6): 67-72.
Wu F, Min H S. A method of real-time 3D semantic map generation ［J］. Computer Engineering and Applications, 2017, 53(6): 67-72.
［6］朱奇光, 王梓巍, 陈颖. 基于图像匹配的移动机器人导航研究［J］. 计量学报, 2017, 38(5): 571-575.
Zhu G Q, Wang Z W, Chen Y. Research on Mobile Robot Navigation Based on Image Matching ［J］. Acta Metrologica Sinica, 2017, 38(5): 571-575.
［7］席志红, 韩双全, 王洪旭. 基于语义分割的室内动态场景同步定位与语义建图［J］. 计算机应用, 2019, 39(10): 2847-2851.
Xi Z H, Han S Q, Wang H X. Synchronous positioning and semantic mapping of indoor dynamic scenes based on semantic segmentation ［J］. Journal of Computer Applications, 2019, 39(10): 2847-2851.
［8］熊汉江, 郑先伟, 丁友丽, 等. 基于2D-3D语义传递的室内三维点云模型语义分割［J］. 武汉大学学报(信息科学版), 2018, 43(12): 2303-2309.
Xiong H J, Zheng X W, Ding Y L, et al. Semantic segmentation of indoor 3D point cloud model based on 2D-3D semantic transfer ［J］. Geomatics and Information Science of Wuhan University, 2018, 43(12): 2303-2309.
［9］Wang Li, Li Ruifeng, Sun Jingwen, et al. Multi-View Fusion-Based 3D Object Detection for Robot Indoor Scene Perception［J］. Sensors (Basel, Switzerland), 2019, 19(19): 4092.
［10］Yu C, Wang J, Peng C, et al. Bisenet: Bilateral segmentation network for real-time semantic segmentation［C］//Proceedings of the European Conference on Computer Vision (ECCV), 2018: 325-341.
［11］Chollet F. Xception: Deep learning with depthwise separable convolutions［C］// Proceedings of the IEEE conference on computer vision and pattern recognition, 2017: 1251-1258.
［12］华静, 陈亮, 李子印. 基于多尺度特征融合的图像语义分割［J］. 中国计量大学学报, 2019, 30(3): 323-330.
Hua J, Chen L, Li Z Y. Image semantic segmentation based on multi-scale feature fusion ［J］. Journal of China University of Metrology, 2019, 30(3): 323-330.
［13］钟海军, 胡步发. 基于高层特征融合的图像语义分割［J］. 机械制造与自动化, 2019(3): 178-181.
Zhong H J, Hu B F, Image semantic segmentation based on high level features fusion［J］. Machinery Manufacturing & Automation, 2019(3): 178-181.
［14］Chen L C, Papandreo G, Kokkinosi I, et al. DeepLab: Semanticimagesegmentation with deep convolutional nets, atrous convolution, and fully connected CRFs［J］. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2016, 40(4): 834-848.
［15］胡涛, 李卫华, 秦先祥. 图像语义分割方法综述［J］. 测控技术, 2019, 38(7): 8-12.
Hu T, Li W H, Qin X X. A review on image sematic segmentation［J］. Measurement & Control Technology, 2019, 38(7): 8-12.
2021, Vol. 42 
