基于聚类的图像分割方法 ·目标应具有相似的亮度、颜色、纹理等 GS04 NC05 TP09 QS09 GCal0 GCblo SLIC Fig.1.Images segmented using SLIC into superpixels of Fig-7.Visual comparson of superplxels produced by various methods.The average superpbel size in the upper lett of each image is 100 pixels and is 300 in the lower size 64,256,and 1,024 pixels (approximately). R.Achanta,A.Shaji,K.Smith,A.Lucchi,P.Fua,and S.Susstrunk,SLIC Superpixels Compared to State-of-the-art Superpixel Methods.IEEE Transactions on Pattern Analysis and Machine Intelligence,2012.34(11):p.2274-2282. Hangzhou Dianzi University抗州电子科技大学 School of Computer Science and Technology计算机学院周文晖
Hangzhou Dianzi University 杭州电子科技大学 School of Computer Science and Technology 计算机学院 周文晖 基于聚类的图像分割方法 Fig. 1. Images segmented using SLIC into superpixels of size 64, 256, and 1,024 pixels (approximately). R. Achanta, A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Süsstrunk, SLIC Superpixels Compared to State-of-the-art Superpixel Methods. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012. 34(11): p. 2274-2282. •目标应具有相似的亮度、颜色、纹理等
基于轮廓和边界的图像分割方法 CSE 803 Fall 2008 Stockman Iaug3 hou Dianzi乙niversity杭州电子科技大学 School of Computer Science and Technology计算机学院周文晖
Hangzhou Dianzi University 杭州电子科技大学 School of Computer Science and Technology 计算机学院 周文晖 基于轮廓和边界的图像分割方法 CSE 803 Fall 2008 Stockman
区域融合和区域增长方法 ·目标区域应具有连续性和相邻性等 把图象看成3-D地形的表示，即2-D的地基（对应图像空间）加上第3维的高度（对应图像灰度） 图像的梯度图也可以视为3D地形，图像梯度图3D地形中的峰岭对应目标的边界 Object1+—Object22 分水岭线 蓄水盆地 盆地 http://en.wikipedia.org/wiki/Watershed_(image_processing) Iaug3 hou Dianzi University抗州电子科技大学 School of Computer Science and Technology计算机学院周文晖
Hangzhou Dianzi University 杭州电子科技大学 School of Computer Science and Technology 计算机学院 周文晖 区域融合和区域增长方法 •目标区域应具有连续性和相邻性等 http://en.wikipedia.org/wiki/Watershed_(image_processing) 把图象看成3-D地形的表示，即2-D的地基（对应图像空间）加上第3维的高度（对应图像灰度） 图像的梯度图也可以视为3D地形，图像梯度图3D地形中的峰岭对应目标的边界
分水岭分割实例 Initiat inage Topographic surface 原始图 梯度图 梯度图上的分水岭原图上的分水岭 Initial inage Finat uatersheds 原始图 阈值分割 分水岭 叠加轮廓 Iaug3 hou Dianzi University抗州电子科技大学 School of Computer Science and Technology计算机学院周文晖
Hangzhou Dianzi University 杭州电子科技大学 School of Computer Science and Technology 计算机学院 周文晖 分水岭分割实例
主动轮廓方法 Given:initial contour(model)near desirable object Goal:evolve the contour to fit exact object boundary 参数化主动轮廓：Snakes Snakes:Active contour models M Kass,A Witkin,D Terzopoulos-International journal of computer vision,1988-Springer A snake is an energy-minimizing spline guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges.Snakes are active contour models:they lock onto nearby edges,localizing them accurately.Scale-space... ☆99被引用次数：21696相关文章所有47个版本0 几何主动轮廓：水平集方法(level set)) >From Snake to Level Set >Classical Level Set Model [Kass,Witkin, Tracking Heart Ventricles Terzopoulos 1988] Hangzhou Dianzi Universit抗州电子科技大学 School of Computer Science and Technolog)计算机学院周文晖
Hangzhou Dianzi University 杭州电子科技大学 School of Computer Science and Technology 计算机学院 周文晖 主动轮廓方法 Given: initial contour (model) near desirable object Goal: evolve the contour to fit exact object boundary [Kass, Witkin, Terzopoulos 1988] Tracking Heart Ventricles 参数化主动轮廓 ：Snakes 几何主动轮廓：水平集方法 (level set) From Snake to Level Set Classical Level Set Model