Development of an Embedded Palm Vein Imaging Prototype

This paper shares one of the available options in developing an embedded palm vein imaging prototype. The prototype was developed by the Raspberry Pi SBC to promote portability of the embedded system. With an integrated illumination circuit utilizing two near infrared (NIR) peak wavelengths of 850 nm and 870 nm, this paper explores the ability of the prototype to capture palm vein pattern information. The prototype program, and image analysis were executed by Python language environment and OpenCV module binding. The captured palm images were compared with palm image datasets from the Chinese Academy of Sciences’ Institute of Automation (CASIA) and the Hong Kong Polytechnic University (PolyU). The comparison was done in terms of observation of the image recorded and palm vein pattern revealed, and also through image assessment metrics. Results show that palm images captured by the prototype has the ability to record vein pattern information in the image with pixel-by-pixel similarity rate of 96.54% (median) for the extracted vein pattern, compared to the CASIA (median: 96.07%) and PolyU (median: 90.99%) datasets. As such, the developed prototype can be enhanced its usage not only for biometric acquisition, but also for medical purpose. Â