A Method for Contactless Palm ROI Extraction Ahmed S. ELSayed 1 , Hala M. Ebeid 2 , Mohamed Roushdy 1 and Zaki T. Fayed 1 1 Computer Science Dept., 2 Scientific Computing Deptarment Faculty of Computers and Information Sciences, Ain Shams University Cairo, Egypt ahmed_salah@cis.asu.edu.eg, halam@cis.asu.edu.eg, mroushdy@cis.asu.edu.eg, ztfayed@hotmail.com Abstract—Palmprint can be extracted from a hand using a low-cost webcam in a contactless manner. Using a webcam makes the enrollment process fast and convenient for users. Being contactless solve the hygiene issue and avoid copying the latent prints from sensor's surface. However, a number of challenges arise in such environment; geometric transformations, the existence of finger rings, hand accessories, and other false objects. This paper proposes a palm ROI extraction method that is robust to these challenges. The method is based on blob analysis, morphological and geometrical operations without a need to pre-train or parameter adjustment. It's tested on three available hand databases that cover these challenges; namely, Sfax, IITD and PolyU 3D/2D. The palm ROI is considered to be wrongly extracted if it contains part of the background. The method achieves an extraction error of 0%, 0.27% and 0.26% for the three DBs, respectively. Applying a massive rotation and scaling tests leads to a minor increase in the extraction errors by 0.24%, 0.35% and 0.84% for the three DBs, respectively. Keywords—palmprint; palm ROI extraction; blob analysis; I. INTRODUCTION Among the various biometric traits, the human hand is the oldest, and perhaps the most successful form of biometric technology. Three traits can be extracted from the hand using the optical camera in a contactless manner: hand geometry, palm print, and knuckle print. These hand traits are stable and reliable. Once a person has reached adulthood, the hand structure and configuration remain relatively stable throughout the person's life [1]. Using a single camera to capture multimodal traits can facilitate the enrollment process and make it more convenient for users. One advantage of using such camera is its low cost, compact size and widespread availability in laptops and mobiles. This allows the system to be used by a large number of users in wide-range of applications (e.g. personal accounts over the web and attendance systems). Being contactless give the system more advantages [2]. First, addressing the hygiene issue in which people have to place their hands on the same sensor where others have also placed theirs. Second, avoid the possibility of copying the latent hand prints that remain on the sensor's surface. Finally, avoid the problem of contaminating the device surface in harsh, dirty, and outdoor environments. A typical hand biometric system consists of five steps; hand acquisition, hand segmentation, region-of-interest (ROI) extraction, feature extraction and matching. After segmenting the hand from the background, the extraction of the hand geometric points and the palm ROI is considered an important step for hand biometric systems. However, many challenges arise from using the contactless camera in an unconstrained and guideless environment which highly affect this step. First, geometric transformations can appear due to a different distance and perspective between the hand and the camera. Second, the existence of finger rings and other hand accessories can cause finger/hand holes in the segmented hand, especially when using skin-based segmentation methods. Third, other false-detected objects may appear in the image of the segmented hand. This paper aims to propose a palm ROI extraction method that copes with the above challenges without a need to pre- train or parameters adjustment. The rest of this paper is as follows: Section 2 contains a literature survey. Section 3 contains the detailed description of the proposed method. DBs and evaluation methodology are described in Section 4. Experimental results and discussion appear in Section 5. Conclusion and future works come in Section 6. II. LITERATURE SURVEY AND RELATED WORKS Many palm ROI extraction methods exist in the literature that tried to cope with challenges of using the contactless camera in an unconstrained environment. Examples of such challenges are hand scaling, hand rotation, partially segmented hand and the existence of finger holes and connected fingers. Y. Feng et al. [3] proposed a real-time palm ROI extraction method based on enhanced version of the competitive hand valley detection (CHVD) algorithm [2]. First, a course-to-fine approach is applied to detect the key points on the hand. Then, the shape context descriptor is used to verify these key points. The method can cope with partially segmented hand and connected fingers. It's robust to scale variation. However, the method doesn't cope with finger holes. Hand left/right determination, fingers classification and valley points detection are provided. A training phase is required for validating the key points. An automatic accuracy measure is applied on a proprietary DB and achieves a 93.8% correct extraction rate. L. Mestetskiy et al. [4] proposed a method for extracting the hand valleys and tips points. The method is based on the analysis of continuous skeletons of binary images. The approach includes polygonal approximation of binary image, skeleton construction, and regularization by pruning. The 978-1-5090-3267-9/16/$31.00 ©2016 IEEE 193