Texture based Animal Segmentation in Aerial Videos

Abstract

Animal detection in aerial videos is a challenging problem due to the complex nature of the scenes involved as well as the natural ability of the animals to camouflage their environment. To assist with the detection and classification of animals for the purpose of nature conservation management, texture analysis is applied to aerial videos of wildlife scenes to segment the environment from the animals. To perform automatic wildlife surveying and animal monitoring, it is proposed to use GLCM texture segmentation to reduce the search area for animals in the aerial videos. Using the texture in the scene, the issues of a moving background and unpredictable state of the animal are avoided. The method presented is well suited to implementation on a UAV as it is easily parallelizable.