MAPML_Segmentation_Program
所属分类:matlab编程
开发工具:Visual C++
文件大小:72KB
下载次数:18
上传日期:2009-08-22 07:25:17
上 传 者:
keller
说明: Image Segmentation by MAP-ML Estimations
文件列表:
GCSeg.mexw32 (32768, 2008-07-16)
kmeansColor.p (1246, 2008-07-16)
featuresExtract.p (15443, 2008-07-16)
MAPMLSeg.p (10411, 2008-07-16)
14037.jpg (46483, 2003-11-01)
This is the program of image segmentation based on the following work:
S. Chen, L. Cao, J. Liu, and X. Tang, "Image Segmentation by MAP-ML Estimations", submitted to IEEE Transactions on Image Processing.
Note that this program was only tested on 32-bit PCs installed with Windows XP or Windows Server 2003. Since the K-means algorithm used in this program is from Matlab Statistics Toolbox, the program runs slower than when a faster K-means algorithm is used which can be obtained at http://vision.ucsd.edu/~pdollar/toolbox/doc/. The program was tested using Matlab R14 and the image processing Matlab Toolbox is required to run it.
This folder includes three Matlab P-code files and a Matlab Mex-file. MAPMLSeg.p is the main function and calls the other two P-code files and the Mex-file.
MAPMLSeg.p: the main function to perform the segmentation.
Usage: MAPMLSeg(sourcepath, filename, targetpath, CluNum, SmoothFac, IteNum)
Input:
sourcepath: the directory containing the image to be processed.
filename: the filename of the image.
targetpath: the directory to save the segmentation results (all results after different iterations).
CluNum: the cluster number in the initial K-means algorithm (default value is 10).
SmoothFac: the smoothness factor (default value is 4000).
IteNum: the maximum iteration number (default value is 5).
An Example:
% segment an image in the current folder
sourcepath = cd;
filename = '14037.jpg';
targetpath = [sourcepath '\result'];
MAPMLSeg(sourcepath, filename, targetpath, [], [], []); % [] represents the default value.
近期下载者:
相关文件:
收藏者: