DeepLabV3_DHC:城市无人机遥感图像语义分割
DeepLabV3_DHC:Semantic Segmentation of Urban Unmanned Aerial Vehicle Remote Sensing Image
孙国文 1罗小波 1张坤强2
作者信息
- 1. 重庆邮电大学计算机科学与技术学院重庆空间大数据智能技术工程研究中心,重庆 400065
- 2. 昆明理工大学信息工程与自动化学院,云南 昆明 650500
- 折叠
摘要
高分辨率无人机遥感图像具有极为丰富的语义和地物特征,在语义分割中容易出现目标分割不全、边缘信息缺失、分割精度不足等问题.为了解决上述问题,基于DeepLabV3_plus模型提出改进的DeepLabV3_DHC.首先,利用多种主干网络进行下采样,采集图像的低级特征和高级特征.其次,将原模型的atrous spatial pyramid pooling(ASPP)全部替换成深度可分离混合空洞卷积,同时添加自适应系数,减弱网格效应.之后,抛弃传统上采样的双线性插值法,替换为可学习的密集上采样卷积.最后,在低级特征中串联注意机制.选用多种主干网络进行实验,数据集选用四川省隆昌市地区的部分图像,采用平均交并比和类别平均像素准确率作为评价指标.实验结果表明:所提方法不仅具有较高的分割精度,而且减少了计算量和参数量.
Abstract
High-resolution unmanned aerial vehicle remote sensing images have extremely rich semantic and ground feature features,which are prone to problems such as incomplete target segmentation,missing edge information,and insufficient segmentation accuracy in semantic segmentation.To solve the above problems,based on DeepLabV3_plus model,an improved DeepLabV3_DHC is proposed.First of all,multiple backbone networks are used for down-sampling to collect low-level and high-level features of the image.Second,the atrous spatial pyramid pooling(ASPP)of the original model is replaced by a depthwise separable hybrid dilated convolution,and an adaptive coefficient is added to weaken the mesh effect.After that,the traditional sampling bilinear interpolation method is abandoned and replaced by the learnable dense upsampling convolution.Finally,cascade attention mechanism in low-level features.In this paper,a variety of backbone networks are selected for the experiment,and some images of Longchang City,Sichuan Province are selected for the dataset.The evaluation index uses the average intersection and combination ratio and the average pixel accuracy of the category as the reference basis.The experimental results show that the method in this paper not only has higher segmentation accuracy,but also reduces the amount of computation and parameters.
关键词
城市无人机遥感图像/语义分割/深度可分离混合空洞卷积/密集上采样/注意力机制/网格效应Key words
urban unmanned aerial vehicle remote sensing image/semantic segmentation/depthwise separable hybrid dilated convolution/dense upsampling convolution/attention mechanism/grid effect引用本文复制引用
基金项目
国家重点研发计划政府间国际科技创新合作项目(2021YFE0194700)
重庆市高技术产业重大产业技术研发项目(D2018-82)
重庆市教委重点合作项目(HZ2021008)
出版年
2024
NETL
NSTL
万方数据