A Low-Cost and Portable Subcutaneous Vein Detection System Using a Combination of Segmentation Methods

Vein detection forms a crucial part in medical diagnostics as it personalizes the health care given to an individual. But the problems related to it cause discomfort to patients. This paper proposes a IR imaging-based vein detection and projection device that is both portable, low-cost and contactless, thus ensuring complete functionality and widespread use even during the COVID-19 Pandemic. The image obtained using an IR camera is pre-processed by enhancing the contrast using Contrast Limited Adaptive Histogram Equalization (CLAHE) and then denoised using Median and Gaussian filtering. Two different pipelines have been tested for image segmentation. The first method employs a combination of local thresholding and Otsu binarization while the other employs a combination of Niblack and Otsu binarization with morphological transformations. It is observed that the first pipeline gave better results in terms of eliminating false positives and negatives in the final image.