Fuzzy and rule-based image convolution

摘要: Abstract An image can be sharpened by high pass filtering of its discrete Fourier transform or an equivalent convolution processing in the spatial domain. Edge detection is a form drastic sharpening with conversion to black and white. The approach taken here edge detection, as well smoothing other use ‘smart’ mask that makes rule-based decisions. Upon employing gains rule consequents, we achieve type fuzzy output pixels. We also enlarge images using our new interpolation obtain higher resolution, detect edges rules then reduce original size better than merely applying image. provide examples directly rules. both without enlargement–reduction process.