The aim of this project is to create a device capable of visualising the human veins in a non-invasive manner. Venipuncture is the medical procedure for drawing blood from the veins. Venipuncture requires the vein to be properly sighted before the drawing of blood, the sighting requires the veins to be first brought into view by building up pressure with the help of a tourniquet. This requires trained professionals to achieve ideal results, any failure from the part of the professional can result in serious side effects like infection or injury. Technology can be put to use to solve this issue, by using IR imaging veins in the body can be precisely visualised. Research findings show that skin readily absorbs light in the NIR spectrum while the veins don’t, this effect is put into use to obtain a clear picture of veins which may be upto 10mm deep. This is especially useful in patients where a vein cannot be found for venepuncture, this is very common in children and older people. A vein imaging device can help the venepuncture to be done with a 100% success rate taking only a fraction of the time otherwise required. This project makes use of Near-IR LEDs to illuminate the area of the skin to be visualised and uses an IR camera to analyse the backscattering from the skin to visualise the veins. The obtained image is then put through a series of image manipulation algorithms to further augment the quality of the visuals and obtain a clearer picture of the veins. This final result is then projected back on to the skin in real-time. Existing vein viewers in the market cost upwards of lakhs of rupees, our prototype is expected to cost under ₹25,000 and can be made much cheaper if the projection feature is omitted. The system makes use of 850nm IR LEDs to achieve illumination of the skin. The brightness is controlled via software PWM. The IR camera uses a Sony IMX219 sensor with 8 megapixels of resolution. This is interfaced with a Raspberry Pi which takes care of all the image processing and display functions. The processed image is then projected back onto the skin at the area being scanned exactly corresponding to the position of the veins. This is accomplished with the help of a DLP projection module from Texas Instruments. All this is enclosed in an ergonomic highly-portable form factor with rechargeable batteries that can be hot swapped helping keep zero downtime. The image manipulation runs on OpenCV on the Raspberry Pi. The image recognition algorithm will be able to compensate for the different skin complexions between patients with minimal changing of parameters. It is also tasked with the reducing the noises induced from ambient lighting and especially sunlight. Color Grading is applied to the image to improve visibility of the projected image.