Vision Based Automatic Inspection System for Nuts Welded on the Support Hinge

This paper is about the development of vision based automatic inspection system of welded nuts on support hinge used to support the trunk lid of a car. Until now, projection welding and inspecting process of welded nuts was being performed manually and recorded by the operator´s naked eye. So these processes caused the production of poorly-made articles. The conditions of poorly-made articles are divided into the two categories. One is an ommition of nuts, the other is an eccenticity of welded nuts. These articles make trouble to erect between support hinge and the lid of trunk. To improve demerits of this manual operation, automatic inspection system is introduced. 3D mechanical design tool, Pro-Engineer, is used to manage the exact tolerance circumstances and avoid design failures. Simulation is performed to make the complete vision based automatic inspection system before assembly, and this procedure led to the manufacturing cost saving. As the inspection algorithm, template matching algorithm (pattern matching and geometric matching techniques in NI IMAQ VISION) is applied to distinguish the articles of good quality and the poorly-made articles. Main panel of developed vision program is programmed by using LabVIEW of National Instruments Co. Main program contains operating mode and selling mode compensating minimum matching score about the variation of the lighting condition. Test algorithm, lighting system and program showed good results to the designed inspection system and led to the increment of productivity in accordance with the reduction of elimination of poorly-made articles