Subcutaneous vein detection and localization for Cannulation
Program: Biomedical Engineering, Multidisciplinary Design
This project aims to design a NIR imaging system and algorithms to perform subcutaneous vein detection and localization. The NIR imaging system uses CMOS detector and NIR filter and is used to capture images of subcutaneous veins on human forearm at 650 nm, 850 nm and visible light. The images are fused to get vein images with better contrast. After that, image segmentation algorithms like Kmeans and maximum curvature method are performed to detect the location of the vein and mark them on the original image.
Professor Israel Gannot
Flow chart of vein localization process of captured image