激光与光电子学进展, 2020, 57 (16): 161012, 网络出版: 2020-08-05
基于多尺度残差注意力网络的壁画图像超分辨率重建算法 下载: 973次
Mural Image Super Resolution Reconstruction Based on Multi-Scale Residual Attention Network
超分辨率 壁画图像 残差网络 注意力机制 多尺度特征 super-resolution mural image residual network attention mechanism multi-scale feature
摘要
壁画图像具有结构细节丰富,纹理复杂、色彩多变的特点,而基于卷积神经网络的图像超分辨率算法重建的壁画图像存在纹理模糊和边缘锯齿效应的问题。因此,提出了一种基于多尺度残差注意力网络的壁画图像超分辨率重建算法。首先,通过多尺度映射单元,用不同尺度的卷积核直接对低分辨率壁画图像进行特征提取;然后,将融合后的特征图输入残差通道注意力块,使网络从全局信息出发对各个特征图进行权值优化,增强网络模型的深度映射能力;最后,在网络末端引入亚像素卷积层,重新排列像素,得到重建的高分辨率壁画图像。实验结果表明,本算法可以减小重建误差,增强重建壁画图像的边缘及结构信息,使重建的壁画图像纹理细节更丰富。
Abstract
Mural image has the characteristics of rich structural details, complex textures and variable colors, while the mural image reconstructed by image super-resolution algorithm based on convolution neural network has problems of texture blur and edge staircase effect. Therefore, we propose a super-resolution reconstruction algorithm based on multi-scale residual attention network. First, the features of low-resolution mural images with convolution kernels of different scales are extracted directly by multi-scale mapping unit. Then, the fused feature maps are input into the attention block of residual channel , so that the weight of each feature map is optimized from the global information, and the depth mapping ability of the network model is enhanced. Finally, a sub-pixel convolutional layer is introduced at the end of the network to rearrange the pixels to obtain the reconstructed high-resolution mural image. Experimental results show that this algorithm can reduce the reconstruction error, enhance the edge and structure information of the reconstructed mural image, and enrich the texture details of the reconstructed mural image.
徐志刚, 闫娟娟, 朱红蕾. 基于多尺度残差注意力网络的壁画图像超分辨率重建算法[J]. 激光与光电子学进展, 2020, 57(16): 161012. Zhigang Xu, Juanjuan Yan, Honglei Zhu. Mural Image Super Resolution Reconstruction Based on Multi-Scale Residual Attention Network[J]. Laser & Optoelectronics Progress, 2020, 57(16): 161012.