Image deblurring algorithm based on improved generative adversarial networks
Jianguo Shi, Jing’Ping Yin, Yule Huang · 2024
The image deblurring algorithm based on improved generative adversarial network is an image restoration algorithm that uses generative adversarial network. In this algorithm, the generator adopts an improved residual block method, which effectively solved the problem of gradient vanishing and reduced network parameters by combining dilated convolution and residual blocks. At the same time, the loss function of the generator has been improved, increasing the generator’s attention to important information and structures in the image. By performing deconvolution pixel transformation on the output of the generator, the deblurred image is obtained. The experimental results show that the algorithm can effectively remove image blur and improve the quality of repaired images.