Hand vein recognition has gained traction in recent years for personal authentication tasks. The dorsal region of human hands constitutes blood veins along with various other features like knuckles. The hand dorsal veins have the strong potential to present unique patterns, which is helpful for biometrics recognition. Additionally, the subcutaneous nature of these hand dorsal veins makes them less prone to professional wear and tear, hence facilitating more reliable usage. In this paper, we explore thermal imaging to investigate the hand dorsal vein segmentation for biometrics applications. A new hand dorsal veins dataset captured by the thermal camera is generated and will be released in public for future research. Moreover, owing to the individual limitations of visible wavelength and thermal imaging, we have explored the blend of both these acquiring wavelengths. Our experimentations reveal that the combination of both the visible and thermal wavelength seems better for vein map segmentation.