Animal Recognition using Principal Component Analysis (PCA)
In this paper, we have analysed Principal Component Analysis, which is one of the most widely used algorithm for image recognition. The experiment is conducted on data set build by us which contains 90 images of each class. There are 2 classes namely Okapi ( a rare animal which is to be identified) and not recognized. The training dataset is a collection of images of okapi. It has been observed that there are various factors that act as challenges in the process of image recognition like illumination, size, orientation, etc. In recent years, a new view-based approach to image recognition has been developed. Comparison of Eigenface and Fisherface approach by size of training data and by image pose is implemented. Here a class will contain all images of one particular animal that is, okapi. The goal is to implement the automated system for recognition of Okapi using their images and recording the amount of time the system takes to perform recognition. This can be easily scaled to many more classes or animals. The procedure of the image recognition system is as follows. First, features of images are extracted. Second, the classifier is trained on training set of images and models for classes are generated. Finally, these classifications models will be used to predict test images. Here we keep the total number of images constant and divide them into different ratios of training and testing images
Keywords - Eigen Values, Euclidean Distance, PCA, Eigen Faces.