Image Deraining for UAV Using Split Attention Based Recursive Network  

基于分散注意力递归神经网络的无人机去雨算法

作　　者：FENG Yidan DENG Sen WEI Mingqiang 冯一箪;邓森;魏明强(南京航空航天大学计算机科学与技术学院/人工智能学院,中国南京211106)

机构地区：[1]College of Computer Science and Technology/College of Artificial Intelligence,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,P.R.China

出　　处：《Transactions of Nanjing University of Aeronautics and Astronautics》2020年第4期539-549,共11页南京航空航天大学学报（英文版）

基　　金：the Fundation of Graduate Innovation Center in Nanjing University of Aeronautics and Astronautics(No.kfjj20191601).

摘　　要：Images captured in rainy days suffer from noticeable degradation of scene visibility.Unmanned aerial vehicles(UAVs),as important outdoor image acquisition systems,demand a proper rain removal algorithm to improve visual perception quality of captured images as well as the performance of many subsequent computer vision applications.To deal with rain streaks of different sizes and directions,this paper proposes to employ convolutional kernels of different sizes in a multi-path structure.Split attention is leveraged to enable communication across multiscale paths at feature level,which allows adaptive receptive field to tackle complex situations.We incorporate the multi-path convolution and the split attention operation into the basic residual block without increasing the channels of feature maps.Moreover,every block in our network is unfolded four times to compress the network volume without sacrificing the deraining performance.The performance on various benchmark datasets demonstrates that our method outperforms state-of-the-art deraining algorithms in both numerical and qualitative comparisons.户外拍摄的图像常受降雨天气影响,造成图像可见度低等问题。作为一种重要的户外图像采集系统,无人机需要合适的去雨算法来提高雨天所摄图像的视觉效果,并给后续计算机视觉任务提供高质量的初始图像数据。为处理不同尺度和方向的雨纹,本文提出一种利用和融合多尺度卷积的多分支网络结构,并与分散注意力机制结合,实现特征级别的多通道交互,从而利用自适应的感受野以处理复杂背景的恢复问题。本文算法将多尺度分支、分散注意力机制和通道分组机制相结合,在实现多尺度处理的基础上与基本残差块的特征通道数保持一致。另外,为更进一步减少网络模型的参数量,本文算法将每个基本模块递归计算4次,从而在保持去雨效果的同时使模型轻量化。大量实验证明,本文算法在多个数据集上均优于目前主流的传统或基于学习的图像去雨方法。

关 键 词：unmanned aerial vehicle(UAV) deep neural network image deraining recursive computation split attention 

分 类 号：TN925[电子电信—通信与信息系统]