摘要
针对雪天气影响共融机器人视觉系统鲁棒性的问题,提出了一种基于雪模型和深度学习融合的去雪算法。根据雪的成像过程推导了一个简化的雪模型,设计了一个基于该模型的深度去雪网络,该网络由雪花检测子网络和去除子网络串联组成。雪花检测子网络采用了残差学习网络,该网络可以准确地学习雪图像和无雪图像之间的差异。去雪子网络采用了密集连接的U型网络。它一方面利用U型网络保留背景的细节信息,另一方面利用DenseNet将低层特征复用到高层的特点来提高去雪的准确度,将它们结合后缓解了去雪过度导致背景细节丢失和去雪不彻底之间的矛盾。试验证明这种基于雪模型的深度去雪网络能够较好地检测和去除图像中的雪花。
Aiming at the problem that snow weather affects the robustness of the fusion robotic vision system, a snow removal method based on snow model and deep learning is proposed. A simplified snow model is derived based on the snow imaging process,and a deep snow removal network is designed based on this model. The network consists of a snowflakes detection sub-network and a snowflakes removal sub-network. The snowflakes detection sub-network uses a residual learning network to accurately learn the difference between snow images and snow-free images. The desnowing sub-network adopts a densely connected U network. It usually can relieve the contradiction of over-desnowing and under-desnowing by using U-net to preserve image details and feature reuse of Dense Net to accuratelly remove snowflakes. Experiments show that the snow model-based deep networks can effectively detect and remove snowflakes from images.
引文
[1]NARASIMHAN S G,NAYAR S K.Contrast restoration of weather degraded images[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2003,25(6):713-724.
[2]FU Xueyang,HUANG Jiabin,ZENG Delu,et al.Removing rain from single images via a deep detail network[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2017:1715-1723.
[3]ZHANG He,PATEL V M.Density-aware single image de-raining using a multi-stream dense network[C]//IEEEConference on Computer Vision and Pattern Recognition.IEEE Computer Society,2018:695-704.
[4]LUO Yu,XU Yong,JI Hui.Removing rain from a single image via discriminative sparse coding[C]//IEEEInternational Conference on Computer Vision.IEEE,2015:3397-3405.
[5]WANG Yinglong,LIU Shuaicheng,CHEN Chen,et al.A hierarchical approach for rain or snow removing in a single color image[J].IEEE Transactions on Image Processing,2017,26(8):3936-3950.
[6]LI Yu,TAN R T,GUO Xiaojie,et al.Rain streak removal using layer priors[C]//Computer Vision and Pattern Recognition.IEEE,2016:2736-2744.
[7]DING Xinghao,CHEN Liqin,ZHENG Xianhui,et al.Single image rain and snow removal via guided L0smoothing filter[J].Multimedia Tools&Applications,2016,75(5):2697-2712.
[8]REN Weihong,TIAN Jiandong,HAN Zhi,et al.Video desnowing and deraining based on matrix decomposition[C]//IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2017:2838-2847.
[9]CAI Bolun,XU Xiangmin,JIA Kui,et al.DehazeNet:An end-to-end system for single image haze removal[J].IEEE Transactions on Image Processing,2016,25(11):5187-5198.
[10]HE Kaiming,SUN Jian,TANG Xiaoou.Single image haze removal using dark channel prior[J].IEEE Trans.Pattern Anal.Mach.Intell.,2011,33(12):2341-2353.
[11]LI Boyi,PENG Xiulian,WANG Zhangyuang,et al.AOD-net:All-in-one dehazing network[C]//IEEEInternational Conference on Computer Vision.IEEEComputer Society,2017:4780-4788.
[12]XU Jiang,ZHAO Wei,LIU Peng,et al.An improved guidance image based method to remove rain and snow in a single image[J].Computer&Information Science,2012,5(3):49-55.
[13]PEI S C,TSAI Y T,LEE C Y.Removing rain and snow in a single image using saturation and visibility features[C]//IEEE International Conference on Multimedia and Expo Workshops.IEEE,2014:1-6.
[14]ZHENG Xianhui,LIAO Yinghao,GUO Wei,et al.Single-image-based rain and snow removal using multi-guided filter[C]//International Conference on Neural Information Processing.Springer,Berlin,Heidelberg,2013:258-265.
[15]GARG K,NAYAR S K.Detection and removal of rain from videos[C]//Computer Vision and Pattern Recognition,2004.CVPR 2004.Proceedings of the 2004IEEE Computer Society Conference on.IEEE,2004:I-528-I-535.
[16]HE Kaiming,ZHANG Xiangyu,REN Shaoqing,et al.Deep residual learning for image recognition[C]//IEEEConference on Computer Vision and Pattern Recognition.IEEE Computer Society,2016:770-778.
[17]RONNEBERGER O,FISCHER P,BROX T.U-net:Convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention.Springer,Cham,2015:234-241.
[18]HUANG Gao,LIU Zhuang,MAATEN L V D,et al.Densely connected convolutional networks[C]//IEEEConference on Computer Vision and Pattern Recognition.IEEE Computer Society,2017:2261-2269.