工业彩色图像的局部自适应滤波

LOCAL ADAPTIVE FILTERING FOR INDUSTRIAL COLOUR IMAGES

针对工业彩色图像的特点和实时性较高的要求,提出了简单的局部自适应滤波方法。该方法可以按照设定参数自动选择局部灰度平稳区,根据选定领域范围内像素的统计特性来判断中心像素的类型,可以较为准确地区分背景点、脉冲噪声点和边界点。结合邻域大小和类型,设计了不同的滤波模板。该方法不需人为选定待滤波区域,效果较好,且运算时间短,满足实时要求。

In this paper, a simple local adaptive filtering algorithm is proposed to meet the industrial colour images＇ characteristics and demands on real-time. In accordance with the setup parameters, this method can select local gray homogeneous area automatically. It can judge the type of central pixel according to the statistical characteristic of pixels in selected regions. As a result, background points, impulse noise points and boundary points can be distinguished accurately. Various filtering templates are designed based on the adjacent pixels＇ type and block size. Without selecting filtering area manually, this algorithm is able to operate fast with good effectiveness and meets the demand of real time.