A Novel Gaussian Measure Curvelet Based Feature Segmentation and Extraction for Palmprint Images

Abstract

Objectives: An effective feature extraction and segmentation model is employed for palm print images to improve accuracy, computation efficiency and robustness of palm print features. Methods/Analysis: The novel Gaussian Measure Curve let based Feature Segmentation and Extraction (GMC-SE) method is introduced for removal of unwanted execution time by using Edge Based Tangent (EBT) model. In addition, to improve the computation efficiency of features being segmented, competent Gaussian measure is obtained by integrating both local and global palm print features. Findings: Experiment is conducted using Poly U 2D palm print database to measure the effectiveness of the proposed work in terms of execution time ratio, computation efficiency, feature extraction accuracy and robustness in palm print recognition. The proposed scheme GMC-SE method is compared against the existing Fine Ridge Structure Dictionary (FRSD) and Personal Identification using Left and Right Palm Print images (PI-LRPP). As a result, the GMC-SE method improves the computation efficiency by 12% compared to existing FRSD model. Conclusion/Application: An effective feature extraction and segmentation are analyzed for palm print images and experimental results are compared. GMC-SE method for palm print images handled different images in an efficient manner compared to existing works.