Blood Vein Detection System for SmartPhones

Needle infusion into the body for the nurses or medical respondents is one of the most common activities in current clinical activities. However, a considerable number of mistakes are occurring by the medical practitioner because of the vein pattern that is beneath the human skin and sometimes unidentifiable. These inaccuracies can lead to unavoidable bad circumstances for both patients and phlebotomists. Here, a mechanism that can alleviate that is necessary. The major objective of this research is to develop a real-time process for the detection of the vein by the phlebotomists along with an application in the most widely used handheld device (Android Smartphones) to avoid or at least diminish the number of inevitable scenarios for the patients and the nurses in their SOPs. Illuminated infrared light in the blood cells of the vein is absorbed due to the manifestation of the Haemoglobin in blood and the IR non-blocking camera can capture the vein patterns in the IR light spectrum. Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm was used to enhance the pattern of the vein in the Android application developed using OpenCV3. Developed system can detect the veins up to 7mm underneath of human skin in real time with a frame rate of 25fps. This is a far better improvement than commercial systems that can detect veins only below 10mm underneath the skin. Moreover, this system not only focused on needle infusion but also it can be used to indicate the place of bleeding for the clots from the human body strokes, etc. in the upper layer of skin. It can also be used to detect & measure liquids in encapsulated in confined dark bottles, for example, liquid chemical pouring into the bottles in the chemical companies, liquid medicine pouring to bottles, etc.