Animal Classification using Facial Images with Score-Level Fusion

Authors

  • Y. M. S. D. Sastry, B. N. Rao, Manepalli Durga Prasad, Mattaparthi Nandini Bhanusri, Ravuri Devi Gayatri Anusha, Saladi Radha Krishna, Peddireddy Valli Baba

DOI:

https://doi.org/10.17762/msea.v71i4.1119

Abstract

A new area of study in machine vision is a real-world animal biometric system that can find and describe animal life in image and video data. These systems use computer vision to figure out how to put animals into groups. We show a new way to classify animal faces based on the score-level combination of recently popular convolutional neural network (CNN) features and appearance-based descriptor features. This method uses a score-level fusion of two different approaches. One uses CNN, which can automatically extract, learn, and classify features, and the other uses kernel Fisher analysis (KFA) to extract features. The proposed method could also be used to classify images and recognise objects in other ways. The results of the experiments show that automatic feature extraction in CNN is better than other simple feature extraction techniques, both for local and appearance-based features. Also, a combination of CNN and simple features with the right score level can get even better results than using CNN alone. The authors showed that the score-level fusion of CNN-extracted features and the appearance-based KFA method have a positive effect on classification.

Downloads

Published

2022-10-18

How to Cite

Y. M. S. D. Sastry, B. N. Rao, Manepalli Durga Prasad, Mattaparthi Nandini Bhanusri, Ravuri Devi Gayatri Anusha, Saladi Radha Krishna, Peddireddy Valli Baba. (2022). Animal Classification using Facial Images with Score-Level Fusion. Mathematical Statistician and Engineering Applications, 71(4), 5320–5337. https://doi.org/10.17762/msea.v71i4.1119

Issue

Section

Articles