【作者】 钱乐乐； 高隽； 徐小红；

【Author】 Qian Lele Gao Jun Xu Xiaohong (School of Computer and Information Hefei University of Technology Anhui Hefei 230009 China)

【机构】 合肥工业大学计算机与信息学院；

【摘要】 本文探讨了一种基于非负特性的图像稀疏编码(non-negative spare coding,NNSC)算法,它是将非负约束应用于稀疏编码而形成的一种新的图像高阶统计特征表示方法。非负稀疏编码(NNSC)模型模拟了人眼以不同通道(ON-channel和OFF-channel)接收非负数据,符合视觉生理特性。该模型在非负性条件约束下,通过减少编码元素之间的高阶相关性,更好地实现对自然图像高阶统计特征的描述。更多还原

【Abstract】 Non-negative sparse coding discussed in this paper is a new method for the representation of higher order statistic feature of natural images based on the non-negative constraint.Non-negative sparse coding(NNSC) model assumes different channels(ON-channel and OFF-channel) that receive non-negative data separately,in line with the physiological characteristics of human eyes.On the condition of non-negative constraint,more robust higher order statistic feature can be extracted from natural images by the model,which further reduces the high order dependency of different elements of the code.更多还原