真彩色图象边缘检测技术在胃镜图中的应用

The True Color Image Boundary Detection Algorithm Applied to Endoscope Image

目前计算机图象处理技术中，边缘检测已有许多成熟的算法、能较好地完成图象分割的任务。但在医学图象中、有些病变区域的边缘很难确定，人眼判断有很大的主观性。用传统的边缘检测技术不会取得好的检测效果。为此，本文针对胃镜图象中胃溃疡边缘的自动检测提出了一种径向搜索算法、成功地解决了计算机自动边缘检测问题，为病变区域的判别与测量提供了客观依据。

Today, many successful algorithms about computerized boundary detection in the field of image processing. are used in image segmentation, but they are not successful when applied to medical images where borders are often not clearly defined and are sometimes difficult even for the human eyes to detect. In these cases, special algorithms are necessary to achieve the required accuracy. This paper presents a radial search algorithm on luminance information towards the automatic detection of gastric ulcers. The algorithm provides the objective evidence for the location and measurement of the disease area.