A revised feather and down recognition model based on MOAA SVM

So far, feather and down category recognition is often done by man with a microscope, but this method has some disadvantages. So a feather and down category recognition system is proposed in the paper, and then feather and down category recognition can be done by computer automatically. After the image processing and segmentation using GA, the triangle node of two-value image of feather and down is to be recognized with SVM, then the triangle nodes which have been recognized will be matched and the distance between the matched triangle nodes is calculated, in the end, the feather and down category is recognized. After lots of experiments, it is found that the recognition rate is lower than artificial recognition. In order to improving recognition rate, RBF kernel SVM and MOAA SVM are introduced into the recognition system, and a revised feather and down recognition model is put forward. It is shown that it is efficient to feather and down recognition.