my project topic is "face recognition" with svm and I have to implement it . so there are many ways and specially new tricks and new topics in some articles for feature extraction before using svm . but I really want the easiest implementation with Matlab! I'm beginner and I don't know PCA , LDA ، Gabor (also heard some others) what's your suggestion? what kind of feature extraction should I use?
For starters, try detecting faces using vision.CascadeObjectDetector in the Computer Vision System Toolbox.
If you have to roll your own, then take a look at this example, showing how to train a classifier to recognize digits using HOG features and SVM.
Note, that the classifier is only one part of the process. You would need to label a lot of faces in the images. You can do that automatically using vision.CascadeObjectDetector, or manually, using the Training Image Labeler App. Then you need to generate a lot of non-face images. Then you need to train a classifier to tell a face from a non-face. Then you need to write a program that would slide a "window" across an image, and classify the region defined by the window as face or non-face.