Self-Monitoring and Detection of Diabetes with Smart Toilet based on Image Processing and K-Means Technique

Abstract

Diabetes is a disorder due to insulin cannot be produced by the pancreas. Furthermore, it may also occur because the body can not effectively use the insulin produced. Sometimes, blood samples are checked by piercing with a needle on the arm or finger three to four times a day. Taking blood to measure the level of sugar in diabetic patients can lead to infection because diabetics did not produce insulin. The authors therefore developed a smart toilet for the diagnosis of diabetes by the colour of urine. The aim of this tool is to make it easier for someone to monitor their health, particularly diabetes. This tool is use the camera as a colour sensor and use the raspberry pi as the main control tool. The smart toilet is installed with the camera and is designed to take an image to produce a red green blue (RGB) colour model. This RGB model is then clustered using the K-means method. Patients are also able to monitor their health by integrating the Internet of Things (IoT) platform. The experimental work is carried out by 20 sample test objects, the results obtained are of high quality and the tool can be used correctly. Whereas in terms of quantity, the urinary colour detection error is about 5%. It can be observed that developed smart toilet is an option for monitoring diabetic activities and controlling the patients glucose levels.