Research Progress in the Field of Image Completion

Abstract

Image completion technology is a challenging research direction in the field of image restoration. The traditional image completion technology mainly fills the missing areas with missing values based on the information of the unmissed areas of the image. Traditional image completion can well complement images with a small missing area and relatively simple texture structure, but it does not work well for images with large missing areas or complex texture structures. With the continuous development of deep learning, the performance of image restoration has been significantly improved. The image completion method based on deep learning can learn the high-level features of the image, so that the result of the completion is more realistic. This article reviews the image completion technology, introduces the basic principles of typical methods and compares their advantages and disadvantages. Finally, we analyze the future research directions in this field and put forward prospects.