Urinalysis is a crucial diagnostic tool, but current dipstick tests face limitations in accuracy and objectivity. While automated urine test strip readers could address these issues, existing devices are often expensive and inaccessible in resource-limited settings. This study presents the development of a low-cost, machine vision-based device for accurate quantification of ketone and glucose levels from colorimetric urine test strips. Our system employs computer vision algorithms to extract RGB color values from test strip images and maps them to analyte concentrations using calibration curves. The device consists of a Raspberry Pi module, a low-cost webcam, and a custom 3D-printed enclosure with controlled LED illumination, all integrated into a portable unit costing under Rs 10,000. We analyzed 8 urine samples with varying ketone and glucose concentrations, capturing images at 2-minute intervals from 2 to 10 minutes post-dipping. The system achieved 88.7% accuracy compared to gold-standard lab measurements, outperforming visual interpretation. This affordable and portable device has the potential to significantly improve access to reliable urinalysis in resource-limited settings, enabling early detection and better management of conditions like diabetes and kidney disease. By providing objective, quantitative results, it could enhance diagnostic accuracy and facilitate more timely interventions, ultimately leading to improved patient outcomes in underserved communities. This innovative solution addresses a significant barrier in healthcare diagnostics and represents a meaningful advance in point-of-care diagnostics, particularly for low-resource areas.