基于Deeplab-v3+的小目标与边缘增强热图像语义分割网络

A small-targets and edge-enhanced network for thermal image segmentation based on Deeplab-v3+architecture

由于热图像存在无颜色信息,边缘模糊,细节信息较弱等问题,较难获得高质量的图像分割效果.为解决这个问题,在编码-解码(encode-decode)架构的基础上,本文增加了多级像素空间注意模块(multi-level pixel spatial attention module,MPAM)、边缘提取模块(edge extraction module,EEM)和小目标提取模块(tiny target extraction module,TTM).其中,MPAM能使网络充分保留细节的同时捕捉到语义信息,EEM和TTM分别提取具有语义信息的边缘和小目标等细节特征.为提高各类别边缘相交区域像素点和小目标物体的预测精度,设计了专门的损失函数对已获得的边缘和小目标特征进行监督训练,提高各类别边缘相交区域像素点和小目标物体的预测精度.将该方法分别应用于课题组构建的热图像数据集SCUT_SEG、公开的热图像数据集SODA和合成热红外数据集Cityscpae,实验结果表明:本文方法比FCN、PSPNet、Deeplabv3+、MCNet、EC-CNN等5种网络分割算法效果略好,性能提升约2.2个百分点.

Despite much research effort that has been devoted to thermal image segmentation,high quality segmented results cannot be readily obtained due to the absence of color information,blurred edges,and weak details in thermal images.Following this trend,we propose a tiny-target and edge-enhanced network for thermal image segmentation based on Deeplabv3+.We design a multi-level pixel spatial attention module(MPAM),an edge extraction module(EEM),and a tiny target extraction module(TTM),with MPAM that enables the network to make full use of the feature and context information of each layer to effectively recover details at the pixel space.Features of edges and small targets are also modeled in EEM and TTM,respectively.Finally,specialized loss functions are designed to supervise these edge and tiny target features to improve the accuracy of small target and pixels along edges.Experiments on three thermal image datasets including SCUT_SEG,SODA and synthetic thermal infrared Cityscapes show that our method is slightly improved by 2.2 percentage points compared with other state-of-art algorithms in the same scene.