作者: D.-M. Tsai , C.-Y. Hsieh
DOI: 10.1016/S0262-8856(99)00009-8
关键词:
摘要: Abstract In this paper we present a global approach for the automatic inspection of defects in directionally textured surfaces which arise textile fabrics and machined surfaces. The proposed method does not rely on local features textures. It is based image restoration scheme using Fourier transform. line patterns any directional textures spatial domain are removed by detecting high-energy frequency components one-dimensional (1D) Hough transform, setting them to zero, finally back-transforming image. restored image, homogeneous region original will have an approximately uniform gray level, whereas defective be distinctly preserved. A statistical process control therefore used set up limits discriminating between patterns. experiments variety fabrics, natural wood shown effectiveness method.