视觉注意计算模型及其在自然图像压缩中的应用

Visual attention computational model and its application in natural image compression

为了提取自然图像中的主要视觉信息以便更好地对图像进行压缩,提出了一种新的基于感兴趣区域的图像压缩方法.该方法使用一个基于视觉生理和心理物理实验结果的视觉注意计算模型计算图像中的感兴趣区域,并用JPEG算法对感兴趣区域和背景区域采用不同的压缩比进行压缩,对部分图像进行了初步实验.结果表明,用该方法压缩后图像的字节数和每像素比特数等参数均好于JPEG算法;同时压缩后图像在视觉上对比突出了感兴趣区域,有利于对感兴趣区域的观察.

In order to compress natural images more efficiently, a novel region of interest （ROI） based image compression approach was proposed. The first step of the approach was to extract primary visual information by finding out the ROI in images. A saliency-based bottom-up visual attention computational model which was motivated by visual physiological and psychophysical experimental results was used. The second step was encoding and it was based on the joint photographic experts group （JPEG） algorithm. The ROI of the image was compressed with a low compression ratio and the background with a high one. The reconstruction approach of the compressed image was like that of the JPEG algorithm. Preliminary experimental results showed that the approach proposed has higher compression ratio than JPEG algorithm and that the images compressed have perceptually high quality.