hy guys,
I am doing a project on "Moving Human Detection using SVM in FPGA". Can anyone suggest me what kind of feature extraction algorithms can i use for training my classifier? still i dont have any idea to begin my project. Can anyone provide me some guideline?
This as already mentioned in other similar threads:
Each dataset asks for different treatments, there is no general rule when dealing with image classification. You should firstly know deeply your pictures and then decide for the suited approach to minimize the size and enhance it if necessasry. Take a look on Weka tool ( the last version ), which has built-in SVM algorithm so that you can make several experiments before coding.