摘要： 针对对比度低、边缘模糊和噪声干扰强的红外图像，提出一种融合非线性扩散与全变分的形状保持模型。其中，模型保真项用于保持图像形状，保留边缘、细节信息，范数项用于图像去噪。实验结果表明，与同类算法相比，该模型能更有效地去除噪声，保持图像形状，并增强图像对比视觉效果。
关键词:
非线性扩散,
全变分,
形状保持,
图像增强,
图像熵
Abstract: Aiming at infrared image with low contrast, edge blurring and strong noise, this paper presents a mathematical model on shape preserving by combining nonlinear diffusion with total variation, where fidelity term is implemented to preserve image shape, edge and detail information, and norm term is used to denoise image. Experimental result indicates that compared with similar algorithms, the model can eliminate noise, preserve shape and enhance contrast effect.
Key words:
nonlinear diffusion,
total variation,
shape preserving,
image enhancement,
image entropy
中图分类号:
王忠华, 邓鹤, 刘建国, 陈忠. 基于形状保持的图像增强模型[J]. 计算机工程, 2011, 37(21): 194-195,198.
WANG Zhong-Hua, DENG He, LIU Jian-Guo, CHEN Zhong. Image Enhancement Model Based on Shape Preserving[J]. Computer Engineering, 2011, 37(21): 194-195,198.