基于上下文的高阶可导图像阈值去噪方法

Threshold Denoising Method for High Order Derivative Image Based on Context

针对传统的硬软阈值函数不连续以及非自适应的缺陷,分别探讨了具有连续性和高阶可导特性的阈值函数、基于局部上下文自适应的阈值函数的构造,并提出基于局部上下文的连续高阶可导阈值函数的构造,并用于图像去噪,实验结果证明了这些方法的有效性。

Aiming at the defects of discontinuity and non-self-adapting of the classical threshold function, three novel threshold functions are presented. They are one with the characters of continuity and high order derivative, one with local context adaptive and one with local context adaptive. The structure of the continuous and high order derivative threshold function with local context adaptive is investigated. The function is used in image denoising. Experimental results indicate that these novel threshold functions are valid in improving denoising performance.