计算机工程与应用2019，Vol.55Issue(20)：139-144,6.DOI:10.3778/j.issn.1002-8331.1806-0259
结合Inception模型的卷积神经网络图像去噪方法
Inception Model of Deep Convolutional Neural Network for Image Denoising
摘要
Abstract
In order to remove the noise in images more effectively, a deep Convolutional Neural Network(CNN)com-bined with inception model is proposed for image denoising using the integrated image as input and output. Multiple spa-tial scale features are densely extracted through the inception structure to enhance the learning ability of the network. Rec-tified Linear Unit(ReLU)is used as an activate function to avoid the vanishing gradient problem. Batch Normalization (BN)and Residual Learning(RL)are utilized to speed up the training process as well as boost the overall denoising per-formance. The experimental results of three Gaussian noise levels based on the public dataset BSDS300 show that the pro-posed model has a 1.28 dB average Peak Signal to Noise Ratio(PSNR)increase and better visual results while reducing computational complexity and improving convergence rate.关键词
图像去噪/深度卷积神经网络/Inception模型/批量规范化/残差学习Key words
image denoising/deep convolutional neural network/Inception model/batch normalization/residual learning分类
信息技术与安全科学引用本文复制引用
李敏,章国豪,曾建伟,杨晓锋,胡晓敏..结合Inception模型的卷积神经网络图像去噪方法[J].计算机工程与应用,2019,55(20):139-144,6.基金项目
国家自然科学基金(No.61574049,No.61772142) （No.61574049,No.61772142）
广州市珠江科技新星项目(No.201806010059). （No.201806010059）
