A mechatronic vision system for inspection of garnished wall plates

Detecting drawn objects on reflective surfaces is a challenging process due to severe reflectivity of illumination even with the utilization of inclined light sources. Lens distortion, which is a common problem, could jeopardize the accuracy of metrics needed during assessment of ceramic plates to classify its quality. This paper proposes a complete mechatronic vision system covering all aspects of building a robust system for inspecting garnished wall plates.