【摘要】 <正> 1.引言基于视觉特性的图像处理技术越来越受到人们的关注,例如:基于视觉特性的图像编码,图像融合等。主要是因为传统的图像处理方法考虑了图像像素的统计特性,利用像素之间的统计冗余达到压缩的目的,或者利用像素的灰度分布实现图像的增强等。实际上,对图像处理的最终结果进行评价的主体是人,统计特性好的图像其视觉效果未必理想。因此,在进行图像处理时如果能够在充分利用图像统计特性的同时考虑人眼的视觉特性,对改善图像处理结果的主观质量和客观质量无疑是可取的。本文正是基于这个观点,提出了一更多还原

【Abstract】 In many various image-debluring methods, Visual Characteristics are not consides. This paper proposes an image-debluring method based on the frequency and direction characteristics of human visual system by weighting high frequency subband of blured image’s wavelet transformation, the enhanced blured image in high frequecy is obtained through wavelet anti.transformation ,and the blind deblured methed is used to restore image from the enhanced blured image,the restored image is well improved in visual quality.更多还原