This paper presents a method to improve the real time Near-Infrared (NIR) subcutaneous veins localization by using the image processing technique based segmentation and fusion. The goal of this paper is to enhance the extraction of veins from NIR image through a series of image processing stages. In these image processing stages, input image is first partitioned to a new set of intensity band via the segmentation technique and they are re-combined through the fusion procedure in order to obtain the accurate information for next analysis procedure. After that, the procedure of segmentation and fusion is followed by the procedure of filtering and aggregation to produce an optimum image in the result. Lastly, the output from this veins extraction enhancement model is able to provide an optimum image for a better clinical diagnose of veins. © 2015 IEEE.