Image inpainting aims to restore the pixel features of damaged parts in incomplete image and plays a key role in many computer vision tasks. Image inpainting technology based on deep learning is a major current research hotspot. To deeply understand related methods and technologies, this article combs and summarizes the latest research status in this field. Firstly, we summarize inpainting methods of different types of neural network structure based on deep learning, then analyze and study important technical improvement mechanisms. In addition, various algorithms are comprehensively reviewed from the aspects of model network structure and restoration methods. And we select some representative image inpainting methods for comparison and analysis. Finally, the current problems of image inpainting are summarized, and the future development trend and research direction are prospected.

图像修复旨在恢复不完整图像中损坏部分的像素特征，在许多计算机视觉任务中起着关键作用。基于深度学习的图像修复技术是当前的主要研究热点。为深入了解相关方法和技术，本文梳理总结了该领域的最新研究现状。首先总结了基于深度学习的不同类型神经网络结构的修复方法，然后分析研究了重要的技术改进机制。此外，还从模型网络结构和恢复方法等方面对各种算法进行了综合评述。并选取了一些具有代表性的图像修复方法进行对比分析。最后总结一下目前图像修复存在的问题，