Affiliation:
1. School of Mechatronic Engineering and Automation Shanghai University Shanghai China
2. Microelectronics Research and Development Center Shanghai University Shanghai China
Abstract
In the field of digital image processing, high dynamic range (HDR) image technology is widely used in digital photography, medical imaging, virtual reality and other fields. However, inconsistent motion areas may occur when shooting dynamic scenes. Therefore, ghost detection needs to be performed on the image sequence. When an image sequence with moving areas is processed using the static multi‐exposure method and fused to generate an HDR image, since the moving area occupies a certain proportion, the moving area in the final image will have overlapping images, unclear details, and reduced image quality. In response to the above phenomenon, this paper proposes a ghost removal fusion algorithm based on enhanced reference images. While ensuring more texture details in the image, it also reduces the fusion time of the image. The fused image has natural colors and can effectively solve the ghost phenomenon that occurs during the processing of HDR images.