Fuzzy based Edge Enhancement using Wavelets and Energy Operator (2005.07305v1)

Abstract: Edge enhancement and preservation of edges to emphasize the features for images is an essential task in computer vision. The conventional operators may cause false edge detection. In this paper, a fuzzy inference system (FIS) is proposed, which combines with Wavelet (WT) based multidirectional and multilevel edge finding technique for getting good edge localization. After applying of WT decomposition on the original image, the coefficients are filtered out from horizontal, vertical, and diagonal directions. A non-linear energy operator is applied to all WT decomposition to corroborate the edge pixels. This energy calculated edge coefficients are given into the FIS for better enhancement. Since WTs allow for separation of the image into frequency bands without affecting spatial locality, the real edge pixels are further emphasized.