综合利用通用霍夫变换与Snake算法对序列图像的分割  

Sequence segmentation algorithm by unifying GHT transform and Snake

作　　者：邓航[1] 余松煜[1] 

机构地区：[1]上海交通大学图像所,上海200030

出　　处：《红外与激光工程》2000年第2期9-11,25,共4页Infrared and Laser Engineering

摘　　要：提出了一种综合算法对图像序列进行分割 :首先根据上一帧图像物体形状信息用霍夫变换确定在当前帧中同一物体的大致轮廓、位置 ,再以此轮廓作为初始值 ,用Snake算法检测出物体的局部形变 ,对于序列的第一帧用手工勾出目标物体大致轮廓。由于通用霍夫变换抗噪声能力强 ,而Snake能准确地找出局部形变物体的边缘 ,综合两种算法的特点能精确地分割出复杂背景下特定的物体。A sequence segmentation algorithm is proposed: first the rough contour and location of an object in the current frame is detected by generalized Hough transform (GHT) based on the shape information of the object in the previous frame, and then with this initial contour the local contour deformation of the object is traced out by Snake, the contour in the first frame can be lined out manually. Due to the resistance of GHT to noise and the ability of Snake to find deformable contour accurately, this algorithm can detect a special object in a complex background with precision. \;

关 键 词：图像分割 霍夫变换 SNAKE算法 计算机视觉 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]