一种提取图像目标边缘的新方法

New method for image target edge extraction

提出一种基于阈值分割技术和形态学相结合的图像目标边缘提取方法,用Top-hat变换后的图像与原始图像相加再减去Bottom-hat变换后的图像以得到最大对比度的图像,继而进行自动图像阈值分割,再用圆形模板执行形态学闭合操作提取边缘,完全去除内部不感兴趣的细节,并保持边缘的连贯性。仿真结果表明,该方法能很好地提取目标边缘,而实际计算量只有传统方法的44.6%～70.2%,且具有较好的抗噪声能力。

A new method to extract target edge based on morphology and area division technology is proposed. First, an image with maximal contrast is achieved by subtracting the image obtained by bottom-hat transform from the image obtained by Top-hat transform plus original image, then a threshold is automatically divided to gain a binary image, and finally morphological operators with a circle model are executed to detect target image edge so as to eliminate inner details and retain continuity. Simulations show that this method can efficiently extract target edges with only 44.6%-70.2% of calculation load as that of traditional method and better performance of noise resistance.