| 摘要 |
<p>An image processing method is used to decide which pixels having a value in some pre-determined range are associated to form an object. The method has the steps of: (i) marking every pixel having the pre-determined property; (ii) scanning across the image and for each marked pixel checking whether any of the adjacent pixels already scanned has a label; and (iii) if none of the adjacent pixels already scanned has a label the marked pixel is given a new label; if the adjacent pixels already scanned include one or more pixels having a label, the marked pixel is given that label and if the adjacent pixels already scanned include pixels having different labels the marked pixel is given one of these labels and the fact that the pixels with these labels are all associated is noted. <IMAGE></p> |
