Automatic Human Gender Identification Using Palmprint
Automatic human gender identification can help in a developing number of applications related to human–computer interaction (HCI), human–robot interaction and surveillances technologies. Besides, it can also assist in human face identification by reducing
- PDF / 334,600 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 55 Downloads / 178 Views
Abstract Automatic human gender identification can help in a developing number of applications related to human–computer interaction (HCI), human–robot interaction and surveillances technologies. Besides, it can also assist in human face identification by reducing the issue of comparing to half of the database. Several biometrics have been used to identify the human gender, but no significant achievements have been reported in the literature. In this study, we have taken palmprint biometrics, because it contains sufficient significant discriminating information like ridges, wrinkles, and principal lines. Based on it, we are going to propose an algorithm for automatic human gender identification. It involves three steps: extraction of ROI, features computation, and classification. Gabor wavelets are employed to extract the palmprint features as they are potential in capturing discriminating textural properties of the underlying image. Its performance is evaluated with simple KNN classifier on publicly available CASIA palmprint Database. The results obtained are quite encouraging with average accuracy of 97.90% with 10 cross validation. Keywords Biometrics KNN classifier
Gender identification Gabor wavelets Palmprints
S. S. Gornale A. Patil (&) Department of Computer Science, Rani Channamma University, Belagavi, India e-mail: [email protected] S. S. Gornale e-mail: [email protected] M. Hangarge R. Pardesi Department of Computer Science, KASCC, Bidar, India e-mail: [email protected] R. Pardesi e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2019 A. K. Luhach et al. (eds.), Smart Computational Strategies: Theoretical and Practical Aspects, https://doi.org/10.1007/978-981-13-6295-8_5
49
50
S. S. Gornale et al.
1 Introduction Gender identification was first perceived as an issue in psychophysical studies with insights on the efforts of understanding human visual processing and identifying key features employed to categorize the gender of an individual [1]. In earlier research, it has seen that there exist greater differences between male and female characteristics which can be further used to improve the performances of recognition applications in surveillance and computer vision [2, 3]. Automatic human gender identification using palmprint will be among the next most popular biometric technology, especially in forensic applications, thanks to its uniqueness and strength. In fact, palmprint data can easily be collected with low-cost devices and minimal cooperation from subjects. Moreover, several palmprint properties can contemplate to identify person gender information. Palmprints are collections of two distinguishable properties, in medical terminology that are known as palmar friction ridges and palmar flexion creases. These structures are permanent, unique, and immutable too for an individual [4]. In this study, we propose a system for automatic gender identification using palmprint. This system would be useful to enhance the accuracy of other biometric identification sys
Data Loading...
