Enhancement on Stain Detection for Automatic Handwashing Audit Vision System

Abstract

Hand hygiene of the health care worker is critical to prevent infectious disease such as airborne diseases, nosocomial infections, and Hepatitis A among patients. Currently, the handwashing audit among hospital staffs are done manually by observation from an expert. There is a need for automation ease the process and accuracy in detection using vision system. This paper focus on the enhancement of the established prototype and the detection system by the additional third template. The prototype is enhanced by size reduction, sturdier material and hands placement base. The vision system uses a robust threshold to detect the discolored stain with an enhancement on stain templates. The detection and recognition of palm in images is a key research topic that has attracted attention owing to an unveiling human perception mechanism. The system accuracy has increased by 40% from previous work by the enhancement which for three cases the system accurately detects stains on washed hands with Glogerm by 71% and unwashed hand with GloGerm by 81%.