Incomplete disinfection can cause serious complications in surgical care. The teaching of effective hand washing is crucial in modern medical training. To support the objective evaluation of hand disinfection, we developed a compact, mobile device, relying on digital imaging and image processing. The hardware consists of a metal case with matte black interior, ultra-violet lighting and a digital camera. Image segmentation and clustering are performed on a regular notebook. The hand washing procedures performed with a soap mixed with UV-reflective powder. This results the skin showing bright under UV light only on the treated (sterile) surfaces. When the surgeon inserts its hands into the box, the camera placed on the top takes an image of the hand for evaluation. The software performs the segmentation and clustering automatically. First, the hand contour is determined from the green intensity channel of the recorded RGB image. Then, the pixels of the green channel belonging to the hand are partitioned to three clusters using a quick, histogram based fuzzy c-means algorithm. The optimal threshold between the intensities of clean and dirty areas is extracted using these clusters, while the final approximated percentage of the clean area is computed using a weighting formula. The main advantage of our device is the ability to obtain objective and comparable result on the quality of hand disinfection. It may find its best use in the clinical education and training.