Automated quality inspection of surface defects on touch panels

Abstract

Capacitive touch panels (CTPs) with advantages of water-proof, stain-proof, scratch-proof, and fast response are widely used in various electronic products built in touch technology functions. It is a difficult inspection task when defects imbedded on surfaces of CTPs with structural textures. This research proposes a Fourier transform-based approach to inspect surface defects of CTPs. When a CTP image with four directional and periodic lines of texture is transformed to Fourier domain, four principal bands with high-energy frequency components crisscross at the center of Fourier spectrum. A multi-crisscross filter is designed to filter out the frequency components of the principal band regions. The filtered image is then transformed back to spatial domain. Finally, the restored image is segmented by a simple threshold method and defects are located. Experimental results show the proposed method achieves a high defect detection rate and a low false alarm rate on defect inspection of touch panels.

A novel algorithm for defect inspection of touch panels

“…Generally, the major energy seemly distribute along four directional lines, i.e. horizontal, vertical, and two diagonal lines, as pointed out in [3,4].…”

“…A C C E P T E D M A N U S C R I P T ACCEPTED MANUSCRIPT 5 To investigate the effectiveness of the frequency domain filtering methods, we apply the method presented in [3,4] to two defect-free touch panels, Product I and Product II, and the results are demonstrated in Fig 2. Fig 2.(a)-(b) shows two sensing circuit images from two panels.…”

“…The common point of these methods is to design a filtering mask in frequency domain to remove background patterns such as circuit parts, and only defects appear in the reconstructed image. In [3], the authors analyzed some touch panels with directional line structures and found that the high-energy distributed along four directional lines, i.e.…”

“…For touch panel inspection, the works presented in the literature still treat the sensing circuits as texture patterns, and apply frequency-based filtering methods to extract defects [2,3]. The common point of these methods is to design a filtering mask in frequency domain to remove background patterns such as circuit parts, and only defects appear in the reconstructed image.…”

“…However, few researches on touch panels have been reported [2]- [3]. A touch panel is a piece of equipment that lets users interact with computing devices such as smartphones, personal computers, cashier machines, and ATM, etc.…”