Research On Video Motion Blurred Image Object Restoration

As one of the current popular topics in the field of digital image processing, image restoration technology has received increasing attention. Based on the current research home and abroad, the advantages and disadvantages of the widely-used image recovery technology are analyzed and compared. Discussions and research over the video-based motion-blurred object recovery technology have been done in correspondance to its advantages and disadvantages. An image restoration system based on sequence of pictures has been developed and implemented to restore the motion-blurred object.First, the model to recover the motion blur through a sequence of video frames is introduced. It utilizes the space and time domain information between video frames to build an effective point spread function and then recovers the image. The complementary information in the fuzzy image sequence is used to enhance the recovery result. Second, the specific steps in the video-sequence-based image restoration are described, as well as the corresponding algorithm that recovers a clear image from image sequence. During the recovery process, video frames are used to reduce the computation and the complexity of recovery platform. In order to reduce the system’s data handling capacity so as to obtain a good recovery result and increase the efficiency, kalman filter is used to track the motion-blurred object while the movement of the target area is used to guide the image restoration process. As the same time, ringing effect during recovery process is alleviated through the use of image enhancement over the recovery result.This scheme is tested in the platform of video-based motion-blurred image recovery system. Through comparison with the resuilts from other image restoration algorithms and the evaluation parameters, the experimental results are analyzed, which shows that video-based motion-blurred image recovery system gives a better performance and obtains a bettter recovery.