Resumen
Palmprint trait has emerged as a means of new and practical biometric recognition system. Palmprint is divided into three regions known as interdigital, hypothenar and thenar. These regions contain a bundle of patterns such as creases, ridges, minutiae and pores that are believed to be unique and distinct in establishing the identity of a person. The process of extracting the palmprint ROI is a crucial and important initial process for personal identification. In obtaining palmprint ROI, there are numerous algorithms that have been proposed by past researchers. However, to the best of our knowledge, all of these developed algorithms only extracted a small part of the palmprint region. Due to this limitation, some important features that are used as an individual identification have been neglected. In this research, a modified algorithm for extracting the palmprint ROI has been proposed. The performance of the proposed algorithm is compared to three other existing algorithms. These four algorithms are then tested on palmprints in order to identify the size of ROI area and the features extracted. The result is encouraging. It shows that the proposed algorithm has successfully extracted a larger ROI compared to existing algorithms. These results are useful in providing prior information for future development of biometric recognition system.