Research on Fast Enhancement of Motion Blurred Image Based on Improved Deep Learning

Abstract

In order to solve the problem that the current vision system produces motion blur during imaging, a fast enhancement method for motion blur images based on improved deep learning is proposed. By detecting the contour of the motion blur image target, the Wiener filtering restoration model is established, combined with the improved deep learning method to decompose the gray tone function, and the BNL-Means algorithm is used to calculate the similarity between high-frequency image blocks to improve the accuracy of spatial image feature extraction. Realize the enhancement of motion blurred images. Compared with the existing methods, it is proved by experiments that the accuracy of the fuzzy kernel estimation of the design method reaches 95.63%, which is higher than the comparison of the three literature methods. The method has a better effect of enhancing the motion blur image and has strong practicability.