PCA-Based Animal Classification System

Missing, disappearing, swapping are fundamental problems encountered especially in pet animals that are very similar. Unfortunately, there are few methods that can overcome such problems. Traditionally, animals are known through their external appearances and patterns. Biometric based systems developed for the identification of animals are scarce and many of these systems are inadequate. In this study, a Principal Component Analysis (PCA) based system was developed for the recognition and classification of different species of animals. Thanks to the application software in the structure of the developed system, it is possible to identify the animals most resembling an animal in the image dataset. Experimental studies on cow, cat, dog, goat and rabbit animal species shows a success rate of 92 % in the first nearest recognition and 83 % in the second nearest recognition. It has been seen that the improving of this developed system can be used in the classification process of different kinds of animals.