一种基于视觉模型的图像编码方法

提出了一种新的图像压缩编码方法，根据视觉的生理学和心理学特性，设计一组与编码图像无关且对人眼有意义的模块（视觉模块），利用这些模块对分块子图作匹配。该方法提供了与视觉相一致的较好的图像质量，压缩比与矢量量化相近。由于采用了简单的匹配算法，大大降低了编码的复杂性，此方法适于实时应用场合。

Image Compression Coding Based on a Visual Model

A new approach to image compression coding is propotqed.Aset of visual patterns visually meaningful to a normal human observer is designed.The patterns are independent of theimages to be coded and are based on the psychological and physiological features of the vi-sion. The patterns are used to match the subimage. The method proposed offers an excellentimage quality consistent with the vision and the conpression ratio is comparable to those offered by vector quantization. With a simple matching algorithm the coding complexity is re-markably reduced and the method is thus suitable for real-time application.