Image-Classification-master
所属分类：图形图像处理
开发工具：matlab
文件大小：10728KB
下载次数：113
上传日期：2014-01-26 12:13:35
上 传 者：
guai111
说明： 图像分类程序，此图像分类采用 SIFT + Kmeans 聚类的方法，然后调用 MLP 对其特征进行分类处理，速度实现比较快，正确率高
(Image classification procedures, the use of this image classification method SIFT+ Kmeans clustering, and then call the MLP classification of its features, faster speed to achieve the correct rate)
文件列表:
AdditionalResources (0, 2014-01-17)
AdditionalResources\SURFmex-v2 (0, 2014-01-17)
AdditionalResources\SURFmex-v2\022_0001.jpg (26846, 2014-01-17)
AdditionalResources\SURFmex-v2\Contents.m (317, 2014-01-17)
AdditionalResources\SURFmex-v2\OpenCV win64.txt (624, 2014-01-17)
AdditionalResources\SURFmex-v2\common (0, 2014-01-17)
AdditionalResources\SURFmex-v2\common\gpusurf (0, 2014-01-17)
AdditionalResources\SURFmex-v2\common\gpusurf\CudaSynchronizedMemory.hpp (13025, 2014-01-17)
AdditionalResources\SURFmex-v2\common\gpusurf\GpuSurfDetector.hpp (9822, 2014-01-17)
AdditionalResources\SURFmex-v2\common\gpusurf\assert_macros.hpp (16340, 2014-01-17)
AdditionalResources\SURFmex-v2\common\gpusurf\gpu_globals.h (5467, 2014-01-17)
AdditionalResources\SURFmex-v2\common\surfoptions.m (616, 2014-01-17)
AdditionalResources\SURFmex-v2\common\surfplot.m (673, 2014-01-17)
AdditionalResources\SURFmex-v2\cppmatrix.h (3180, 2014-01-17)
AdditionalResources\SURFmex-v2\examples (0, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\gpu_test.m (1853, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama (0, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\a.jpg (121140, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\b.jpg (124832, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\c.jpg (118686, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\d.jpg (118387, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\estimate_projective4point.m (596, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\estimate_projective_nonlinear.m (986, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\estimate_projective_ransac.m (3732, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\panorama.m (4623, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\panorama\projective_distance2.m (346, 2014-01-17)
AdditionalResources\SURFmex-v2\examples\short_example.m (1719, 2014-01-17)
AdditionalResources\SURFmex-v2\make.m (978, 2014-01-17)
AdditionalResources\SURFmex-v2\mextiming.h (1822, 2014-01-17)
AdditionalResources\SURFmex-v2\mexutils.h (2741, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32 (0, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\cv210.dll (2079232, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\cxcore210.dll (2200064, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\surfmatch.m (1072, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\surfmatch.mexw32 (14848, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\surfpoints.m (1977, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw32\surfpoints.mexw32 (14336, 2014-01-17)
AdditionalResources\SURFmex-v2\mexw64 (0, 2014-01-17)
... ...
SIFT_classify is the code written for classification of 3 class data.
One can directly run the code by changing the path to get the images and
the path to VLFeats setup. The code directly returns confusion matrix.
SURF_classify is the code written for classification of 25 class data.
One can run the code by changing the path to the images and the path to SURFMEX
which is presented in the additional resources folder. It returns the predicted class
matrix and confusion matrix can be obtained by using
[confusionmt,order] = confusionmat(groundtruth,predictedclassmatrix);
Overall accuracy of the classifier can be simply checked by finding
(numel(find(Groundtruth==predictedclass))/292)*100 (%% since there are 292 test images I woked on)
近期下载者：
相关文件：
收藏者：