改进的LIP偏微分方程图像去噪方法

摘要

Concerning the defects of Logarithmic Image Processing-Total Variation (LIP_TV) denoising model, an improved Partial Differential Equation (PDE) image denoising method based on LIP was proposed.Based on LIP mathematic theory, the new LIP gradient operator was obtained by introducing four directional derivatives in the original one, which can control the diffusion process effectively because it measures image information comprehensively and objectively.The fidelity coefficient was constructed by adopting the noise visibility function based on the structure characteristic of human visual system, which can further preserve the edge details and avoid estimating noise level faetitionsly.The theoretical analysis and experimental results show that the improved method has superiority in the visual effect and objective quality, which can better remove noise and preserve detailed edge features.%针对对数图像处理-全变分(LIP-TV)去噪模型存在的不足,提出一种改进的LIP偏微分方程去噪方法.首先基于LIP数学理论,在LIP梯度算子中,引入四方向导数信息,得到改进的LIP梯度算子以全面客观地度量图像信息,更好地控制扩散过程.然后利用人类视觉系统的结构化特性,用噪声可见度函数构造新的保真项系数,进一步保持了图像的边缘细节并避免了人为估计噪声水平.理论分析和实验结果表明,该改进方法能够更好地去除噪声和保持图像边缘细节特征,在视觉效果和客观评价指标上都明显优于LIP-TV方法.