In this study, an automatic blood vessel searching system (BVSS) is newly developed, which is built in the Health Monitoring System (HMS) and the drug delivery system (DDS) to extract the blood, evaluate the blood sugar level and inject the insulin for the diabetic patients. Main subjects of this BVSS development are 1) a transmittance photo imaging of the finger by using the LED light as a near-infrared light source with peak wave length of 870nm, and 2) an image processing to detect the location of the center of blood vessel. To estimate the depth of blood vessel, the lens focal method was applied to the BVSS. To confirm the possibility of the method, we carried out experiments using blood vessel phantoms that consist of a cylindrical blood region (Teflon tube) and skin tissue (silicone rubber). The blood region was 0.6mm in diameter and filled with human blood. The experimental results demonstrated the estimated depth from each image corresponding to given depth exactly and confirmed the validity of BVSS.