基于统计规律的运动模糊图像恢复技术-计算机技术专业论文.docxVIP

Th The restoration techniques of motion blurred image based on statistical regularities Abstract The restoration techniques of motion blurred image based on statistical regularities Abstract The relative movement between camera imaging sensor and the shooting scene can cause a motion blur of the image, making the image quality cannot meet the actual demand. The direction, distance, acceleration and other factors of the relative movement are randomly, which make it difficult to backstepping the process of motion blur. The restoration of motion blurred images has some difficulties, and is a hot issue in the field of image restoration. Currently there are a lot of methods for the restoration of motion blurred images, such as mechanical image motion compensation methods, optical image motion compensation methods, the electronic image motion compensation methods, integrated image motion compensation methods as well as graphical image motion compensation methods. The graphical image motion compensation method is a method for the post-processing of the existing digital image. If we get enough prior knowledge of the original image, we can then built the mathematical model and fit the model with the degraded image. This method has proven efficient.Another approach can be used when lack of a priori knowledge of the original image. We can build the degradation process model of the image, bring the motion blur image into the model, and get the original image. This method is called the blind restoration. In this paper, we focus on the blind rehabilitation process. In order to get the motion parameters in the recovery process, the motion parameter estimation techniques based on statistical regularities is used to automatically restore the image. First, this paper gives a systemic introduction of motion blur, and describes the basic concepts and background of image restoration. The current research status, as well as some common restoration algorithms are also introduced. The MATL