Defect inspection of disposable glucose test strips using machine vision

This paper presents a machine vision system for defect inspection disposable glucose test strips. The proposed system first lightened and captured image of flawless glucose test strip while offline training. The captured image was processed through image processing technology to establish standard projection data of silver paste and carbon rubber in flawless glucose test strips. Then the measured blood glucose test strips were processed with the same lighting and image processing procedures when online detection. The defects of the measured blood glucose test strips were inspected by the projection data comparison. Experiments show that the proposed system can effectively detect three main defects on glucose test strips, i.e., silver exposing, short circuit of carbon paste electrode and unglued carbon plastic electrode.