基于HSI空间颜色三角形的边缘检测方法  

Edge detection based on the measure of color triangle in HSI

作　　者：赵昭[1] 王菁[1] 李宪玲[1] 

机构地区：[1]曲阜师范大学计算机科学学院,山东日照276826

出　　处：《微型机与应用》2009年第20期47-49,55,共4页Microcomputer & Its Applications

摘　　要：利用像素点的颜色坐标H、S、I构建像素的颜色三角形,计算该三角形的面积和边长。根据面积以及边长的差值确定该像素点是否为彩色图像的边缘点。这种边缘检测方法在一定程度上合理地考虑了个各颜色分量的相关性,将向量空间的计算以自然的方式转换成了标量的计算,在思想上是一种不同于其他算法的新算法。实验证明,与传统的方法相比较,该方法能快速有效地检测出图像的边缘。Construct the color triangle of a pixel using its color coordinates, and consequently calculate the area and the side length of the color triangle. It can determine whether a pixel is the edge point of a color image by computing the difference of area or side length among these color triangles. The idea of this new edge point detecting method is different from other methods, because it takes the correlation of every color component into account reasonably, and converts the computation of a vector space to scalar quantity in a natural way. The experiment result proves that this method can detect color edges fast and effectively and be superior to the traditional ones.

关 键 词：彩色图像 边缘检测 颜色三角形 颜色相关 

分 类 号：TN911.73[电子电信—通信与信息系统]