An improved street scene segmentation method was proposed to address the challenges of inaccurate small-object segmentation and high computational cost in autonomous driving. The method was developed based on the DeepLabV3+ framework with integrated attention mechanisms and multi-scale feature extraction. Firstly, the original Xception backbone was replaced by a pre-trained MobileNetV3 network to reduce computational complexity. Secondly, SPM (strip pooling module) was incorporated into ASPP (atrous spatial pyramid pooling) structure to capture elongated spatial features. To address the limitation of single-scale object detection in strip pooling, MSPM (multi-scale strip pooling module) was designed. Finally, deep and shallow features were fused using CGA (content-guided attention) mechanism to preserve essential features, and the Focal Loss function was employed to alleviate class imbalance in the dataset.Experimental results on the Cityscapes dataset indicate that the proposed model achieves mIoU (mean intersection over union) of 59.9% and mPA (mean pixel accuracy) of 90.9%. These represent improvements of 9.73% in mIoU and 0.3% in mPA over the original DeepLabV3+ model. Moreover, the parameter count is reduced by 21.06×106. Comparative evaluations demonstrate that the proposed approach outperforms mainstream semantic segmentation models, including FCN, SegNet, and PSPNet. The study concludes that the method effectively achieves model lightweighting while improving segmentation accuracy for urban street scenes.
WU Jin, XIE Xing-jun, LUO Jun-hang, SUN Pei-yi.
Semantic Segmentation of Urban Street Scenes Based on Improved DeepLabV3+[J].
Science Technology and Engineering, 2026, 26(18): 7759-7768 DOI:10.12404/j.issn.1671-1815.2505785
LongJ, ShelhamerE, DarrellT. Fully convolutional networks for semantic segmentation[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston: CVRR, 2015: 3431-3440.
[2]
RonnebergerO, FischerP, BroxT. U-Net: convolutional networks for biomedical image segmentation[C]// Medical Image Computing and Computer-Assisted Intervention (MICCAI). Munich: MICCAI, 2015: 234-241.
[3]
ZhaoH, ShiJ, QiX, et al. Pyramid scene parsing network[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu: IEEE, 2017: 2881-2890.
[4]
ChenL C, ZhuY K, PapandreouG, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]// Proceedings of the European Conference on Computer Vision (ECCV). Munich: Springer Verlag, 2018: 801-818.
[5]
ChangY, YanL, ZhongS. Hyper-laplacian regularized unidirectional low-rank tensor recovery for multispectral image denoising[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu: IEEE, 2017: 5901-5909.
[6]
ChenL C, PapandreouG, KokkinosI, et al. Deeplab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[7]
ChenY, Pont-TusetJ, MontesA, et al. Blazingly fast video object segmentation with pixel-wise metric learning[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 1189-1198.
[8]
杨海红. 基于卷积神经网络的城市街景图像语义分割研究[D]. 济南: 山东工商学院, 2023.
[9]
YangHaihong. Research on semantic segmentation of urban street scene Images based on convolutional neural networks[D]. Jinan: Shandong Institute of Business and Technology, 2023.
GuXiangyu, LiuXiaoyi, ZhouRenbin. Road scene semantic segmentation algorithm based on multi-feature fusion[J]. Science Technology and Engineering, 2021, 21(33): 14251-14257.
XiongWei, ZhaoDi, SunPeng, et al. Research on urban street scene semantic segmentation method based on transformer architecture[J]. Opto-Electronics Laser, 2024, 35(12):1240-1249.
MengQingcheng, ZhangFa, XiongShuang, et al. Surface rust image segmentation method for existing steel structures based on improved DeepLabV3+[J]. Journal of Civil and Environmental Engineering (Chinese and English), 2025.DOI:10.11835/j.issn.2096-6717.2025.004.
[24]
LiM, WeiL S. Surface defect segmentation of magnetic tiles based on improved DeepLabV3+[J]. Journal of Electronic Mea-surement and Instrumentation, 2025, 39(1): 50-56.
[25]
HowardA, SandlerM, ChenB, et al. Searching for MobileNetV3[C]// IEEE/CVF International Conference on Computer Vision (ICCV). Seoul: IEEE, 2019: 1314-1324.
WangGuoping, HuBo, ChenZhongsheng, et al. Surface defect visual detection method based on edge computing improved MobileNetV1-SSD[J]. Marine Electric & Electronic Technology, 2024, 44(3): 9-14.
[28]
SandlerM, HowardA, ZhuM, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 4510-4520.
YueYongheng, NingRuihou. Object detection algorithm for intelligent vehicles based on lightweight CenterNet[J]. Journal of South China University of Technology (Natural Science Edition), 2024, 52(8): 45-55.
[33]
WangZ, ChenY, WangF, et al. Improved unet model for brain tumor image segmentation based on ASPP-coordinate attention mechanism[C]// 5th International Conference on Big Data & Artificial Intelligence & Software Engineering (ICBASE). Wenzhou: ICBASE, 2024: 393-397.
[34]
HouQ, ZhangL, ChengM M, et al. Strip pooling: rethinking spatial pooling for scene parsing[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 4003-4012.
ZhangHao, HeLingmin, PanChen. Improved SegFormer network for semantic segmentation of remote sensing images[J]. Computer Engineering and Applications, 2023, 59(24): 248-258.
[37]
ChenZ X, HeZ W, LuZ M. DEA-Net: single image dehazing based on detail-enhanced convolution and content-guided attention[J]. IEEE Transactions on Image Processing, 2024 (33): 1002-1015.
[38]
ZhangY, LiS, YaoT, et al. A drone detection network based on multi-scale feature fusion and content-guided attention mechanism[C]// 8th International Conference on Artificial Intelligence and Big Data (ICAIBD). Chengdu: ICAIBD, 2025: 219-224.
[39]
罗芯. 面向自动驾驶的语义分割与目标检测多任务模型研究[D]. 杭州: 浙江大学, 2023.
[40]
LuoXin. Research on multi-task model of semantic segmentation and object detection for autonomous driving[D]. Hangzhou: Zhejiang University, 2023.
[41]
XiaY, QiuJ. SFC DeepLabV3+: a lightweight grape image segmentation method based on content-guided attention fusion[J]. Computers, Materials & Continua, 2025, 84(2): 2531-2547.