A Smart Urine Analysis Device for Diabetic Patients

Numerous cutting-edge technologies with accessible processing techniques are required in the social health care sectors. This might reduce costs and boost accessibility to address various issues with real-time services. The use of a urine dipstick is a kind of the healthcare services that are employed in real-time. Urinalysis, or the examination of a urine sample, is a potent method for obtaining crucial information for diagnostic reasons in medicine. A urine test strip is frequently employed in these kind of evaluation. In this work, a real-time implantation for image-processing-based urine test strip analysis was created. The outcome is determined by the strip's colour. However, it could result in observed variations in the colour reading. In this study, digital image processing methods such segmentation, thresholding, smoothing, and recognition were also applied. For classification, a common back propagation technique is computed here. The statistical interpretation, including the positive (PPV) and negative (NPV) predictive values, was evaluated using cross-validation findings. The output demonstrates that the suggested technique has a 96% accuracy in classifying the colours on a urine dipstick.