Machine vision in detection of corrosion products on SO₂ exposed ENIG surface and an in situ analysis of the corrosion factors

Corrosion resistivity is an important property in printed circuit boards manufacturing where the electroless nickel immersion gold (ENIG) is a commonly used surface finish. The corrosion resistivity of the surface is tested with a sulphur dioxide (SO₂) test which imitates the influence of atmospheric corrosion. The test result is analyzed manually by vision inspection, which is unsuitable for mass production. The result is also qualitative (failed or passed) and therefore the corrosion severity is impossible to analyze and further mathematical analysis is hard to conduct. In this paper, a new machine vision based corrosion evaluation algorithm is developed for quantitative and objective analyzing of the SO₂ test result. The algorithm produces automatically the same data as the manual method and, in addition, also new continuous corrosion state estimates. The produced data are further used for correlation analysis in which the relations between the parameters of the ENIG process and the corrosion resistivity of the surface are studied and the algorithm is verified.