基于数学形态学的自适应彩色图像滤波

摘要

For the color vector representation and impulse noise value of the color image, an adaptive algorithm, which filters impulse noise of color image,is proposed to solve the problem of the color image filtering. Firstly, morphology is used to detect the impulse noise, then, according to the result, an improved method of vector median filter adjusts the filter window adaptively for filtering impulse noise. Experimental results indicate that the proposed method can suppress impulse noise in color images effectively. It provides a better restoration performance than many other filters used for removing impulse noise from color images.%为了解决彩色图像滤波问题,针对彩色图像的颜色矢量表示形式和彩色图像中的脉冲噪声的数值特征,提出一种滤除彩色图像的脉冲噪声的自适应算法.首先应用数学形态工具对脉冲噪声进行检测,再根据检测结果,用改进的矢量中值滤波方法自适应地调整滤波窗口,以符合人眼视觉特性的颜色相似性度量方法选择颜色距离最接近的样本像素,对脉冲噪声给予有选择的滤除.通过实验及与其它算法比较,结果表明该算法对于彩色图像中的脉冲噪声有较好的滤除效果.