论文部分内容阅读
基于图像特征的分析,本文提出了一种能够最大限度地保持图像细节、且同时滤除图像随机颗粒噪声与结构性噪声的多元平滑算法。该算法优于图像系统通常采用的均值与中值滤波方法,能够使用于实时热成像系统图像处理场合
Based on the analysis of image features, this paper presents a multivariate smoothing algorithm that can preserve image details to the maximum while filtering out random particle noise and structural noise of the image. The algorithm is superior to the average and median filtering methods commonly used in image systems and can be used in image processing of real-time thermal imaging systems