icing operation. Finally, the multi-channelР features are sent into the reconstruction network composed of residual network to generate the final enhancedР image. In addition, we use compound loss function to train the low light image data set. Experimental results showР that multi-channel feature extraction can be achieved better by using this method. Compared with the currentР mainstream image enhancement methods, the network enhancement effect proposed by us is superior to otherР methods in subjective vision and objective evaluation indicators.Р Key words: smooth dilated convolution；channel domain attention；residual network；image enhancement；Р low light image
全文预览
