Animal Detection using Background Subtraction & Blob Detection Technique

Abstract

Animal detection plays an important role in day to day life due to its impact on the human life directly or indirectly. In the area like an airport where the presence of any kind of animal is strictly restricted, animal detection tool can play an important role in such areas. In this work, the performance of different image features and classification algorithms in animal detection application, and design a real-time animal detection system following criteria in terms of accuracy, time and cost of computation is explored. To follow these qualities, detection process is done in two levels. In first level, bulb detection process is used to subtract background from the image and this image is used in second stage for finding the region of the object using regionpropos algorithm. To examine the animal detection system, we created our own dataset, this dataset can be updated according to the application or use. The result of the approach shows that we can successfully detect the animal when it comes in a particular background.