Detection of Avian Pox Disease using K-Means and Svm Classifer

Abstract

This paper discusses about various methods involved in detection of avian pox in the birds using images. Digital images are corrupted while sending and receiving the images because of noisy sensors which degrade the quality of image. Pre-processing becomes an initial and crucial step in image processing to remove the noise and maintain fine details and texture of the image. Pre-processed images can be used for further work. Mean, Median, Weiner, Mean Maximum, Mean Minimum filters are used and performance tests are made using Signal Noise Ratio. Based on the performance test, removal of impulse noise is well done by Median filter and produces the best result when compared to other filters. K-Means clustering and SVM are used for identification of the disease.