期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2015
卷号：8
期号：9
页码：31-44
DOI：10.14257/ijsip.2015.8.9.04
出版社：SERSC
摘要：Subject in the process of image acquisition and the relative games between the camera or camera image fuzzy, it seriously affected the further image feature extraction, target recognition and image analysis. In order to meet the needs of reality, get high quality image, we must be used image restoration techniques for fuzzy image restoration. Based on the detailed study on the basis of the theory of wavelet transform, we confirmed by experiments in motion blurred image of wavelet and decomposition level .The method to identify the fuzzy parameters, under the condition of without noise and with noise on the four classic image restoration algorithm, the simulation experiment using the peak signal-to-noise ratio, mean square error, average gray level gradient method and Laplace operator method of four kinds of image quality evaluation index to evaluate the healing effect and recovery result diagram and its quality evaluation. Through the scene shooting motion blurred image point spread function of parameter identification and image restoration, it demonstrates that the proposed point spread function's parameters identification method is sound and effective
