*Corresponding author, e-mail: sirbrw@yahoo.com Research Article GU J Sci 33 (3): 820-833 (2020) DOI: 10.35378/gujs.605631 Gazi University Journal of Science http://dergipark.gov.tr/gujs Deep Learning-Based Architectures for Recognition of Cow Using Cow Nose Image Pattern Rotimi-Williams BELLO * , Abdullah Zawawi Hj TALIB , Ahmad Sufril Azlan Bin MOHAMED School of Computer Sciences, Universiti Sains Malaysia, 11800, Pulau, Pinang, Malaysia Highlights • Deep learning approach for learning discriminating texture features of cow nose image. • Deep learning-based methods enhance animal biometrics. • Stacked denoising auto-encoder for encoding and decoding extracted features. • Deep belief network for learning and representing features. • Deep belief network has 98.99% accuracy compared to other methods. Article Info Abstract Stacked denoising auto-encoder and deep belief network are proposed as methods of deep learning for cow nose image texture feature extraction, and for learning the extracted features for better representation. While stacked denoising auto-encoder is applied for encoding and decoding of the extracted features, a deep belief network is applied for learning the extracted features and representing the cow nose image in feature space. Stacked denoising auto-encoder and deep belief network help in animal biometrics. Biometrics emanated from computer vision and pattern recognition and it plays an important role in the automated animal registration and identification process. Using the visual attributes of cow, and for the fact that the existing visual feature extraction and representation methods are not capable of handling cow recognition; deep belief network and stacked denoising auto-encoder are proposed. An experiment performed under different conditions of identification indicated that deep belief network outshines other methods with approximately 98.99% accuracy. 4000 cow nose images from an existing database of 400 individual cows contribute to the community of research especially in the animal biometrics for identification of individual cow. Received: 16/08/2019 Accepted: 12/02/2020 Keywords Animal biometrics Deep learning Cow nose image SDAE DBN 1. INTRODUCTION For any successful animal husbandry, it is important to have reliable, affordable, scalable and effective livestock management for significant animal performance. Herders in the country face a lot of challenges such as cow rustling due to the nomadic system of grazing which has greatly reduced productivity and profit. The manual identification approach for cow recognition for a long time had posed a difficulty for herders and animal husbandry community in monitoring cows, and the existing cow tracking methods are not accurate enough such as in the case where some external factors such as background patches might distort the images and contribute to the increasing difficulty of the detection process [1, 2]. Hence, there is a need for proper monitoring of cow using recent methods for reliable tracking, recognition, and identification of individual cow in herds. What the animal looks like as a consequence of the interaction of its genotype and the environment is made up of an organism’s observable structural features [3, 4]. Though, the manual framework methods of identification provide traditional methods for individual cow identification in the herd but, they are incapable to provide satisfactorily, the security level for herders and cow breeders in monitoring cow all over the world. Biometrics applications in monitoring animals emanated from computer vision and pattern recognition which are branches of artificial intelligence [3, 5].