金刚石工具表面磨粒的轮廓提取

A Contour Extraction of Abrasive Grain in Diamond Tools

为了实现金刚石工具的机器视觉检测,本文先对金刚石工具表面的磨粒在显微镜下进行图像采集,通过聚焦融合方法生成金刚石颗粒的融合图像与其高度矩阵,然后应用动态轮廓模型(Snake)对金刚石的融合图像进行了分割,结合其高度索引图像生成了适合分割的新的图像能量函数,并采用在多分辨率下用遗传算法进行动态轮廓模型的演化,最终得到分割图像。

In order to detecting the diamond tools by computer vision technology, the fused image of abrasive grain and the height matrix image have been gotten, according to the principle of depth from focus. And then active contour model have been applied to segment the image of grain. By combining the height matrix, we got a new kind of image energy matrix suitable for image segmenttation. In multi resolution, the active contour model has been calculated by genetic algorithm, and the contour of the grain can be gotten.