A Robust Texture Based Ear and Palmprint Recognition Using Histogram of Oriented Gradients
K. Naga Prakash1, Parimi Hema2, K. Prasanthi Jasmine3
1Dr. K. Naga Prakash, Professor, ECE Department, Gudlavalleru Engineering College, Gudlavalleru, India.
2Parimi Hema, M. Tech, ECE Department, Gudlavalleru Engineering College, Gudlavalleru, India.
3Dr. K. Prasanthi Jasmine Professor, ECE Department, Andhra Loyola Institute of Engineering and Technology, Vijayawada, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP:1964-1971 | Volume-8 Issue-6, August 2019. | Retrieval Number: F8558088619 /2019©BEIESP | DOI: 10.35940/ijeat.F8558.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Recent research in the surface-based ear and palm print recognition additionally shows that ear identification and palm print identification. The surface-based ear and palm print recognition are strong against sign corruption and encoding antiques. Based on these discoveries, further research and look at the comparison of surface descriptors for ear and palm print recognition and try to investigate potential outcomes to supplement surface descriptors with depth data. The proposed Multimodal ear and palm print Biometric Recognition work is based on the feature level fusion. Based on the ear images and palm print images from noticeable brightness as well as profundity records, we remove surface with outside labels starting complete contour images. In this paper, think about the recognition performance of choose strategies for describing the surface structure, which is Local Binary Pattern (LBP), Weber Local Descriptor (WLD), Histogram of oriented gradients (HOG), and Binarised Statistical Image Features (BSIF). The broad test examination dependent scheduled target IIT Delhi-2 ear and IIT Delhi palm print records affirmed to facilitate and expected multimodal biometric framework can build recognition rates contrasted and that delivered by single-modular for example, Unimodal biometrics. The proposed method Histogram of Oriented Gradients (HOG) achieving a recognition rate of 124%
Keywords: Binarised Statistical Image Features, Histograms of Oriented Gradients, Local Binary Patterns, Weber Local Descriptor, And Biometric Recognition.