作者: L. Aporius , M. Pfitzer , A. Loos
DOI:
关键词:
摘要: In the ongoing biodiversity crisis many species, particularly primates like chimpanzees or gorillas, are threatened. Therefore, autonomous monitoring techniques become more and important to protect remaining populations. However, manual annotation of images video sequences is not feasible for such a huge amount data. Consequently, there high demand automated analytical routine procedures. Recently, computer vision animal detection identification applied overcome this issue. paper we compare several state-of-the-art algorithms human face recognition very new field primate photo identification. Besides common Eigenfaces, Fisherfaces, Laplacianfaces as well sophisticated approaches Tensor Subspace Analysis Volterrafaces, also use concept using randomly generated projection matrix in conjunction with classifier based on sparse representation. Our experimental results show that Sparse Representation Classifier outperforms all other algorithms.