Urine Color Analysis Using Random Forest Algorithm

Urine color is an indicator of health status, and any changes in urine color might be a symptom of a significant underlying ailment. To get better assistance for urine color detection and to overcome impaired color discrimination by the naked eye, a urine color automatic identification has been developed by using image processing and machine learning based on the Random Forest (RF) algorithm. The proposed system uses a web camera to capture images, analyze them, and then compare them with the dataset learned by using the RF algorithm to detect and classify the color of urine and present the results via Graphical User Interface (GUI). In addition, the proposed system can send the results to the cell phone of the care provider by using an Arduino microcontroller and GSM module. The results showed that the proposed system has high accuracy in detecting urine color under different light conditions, low cost, and it is easy to implement.