ridgeletlike
Ridgeletlike is an umbrella term used in signal processing and computer vision to describe methods, representations, or models that resemble ridgelets in their focus on line- or ridge-like singularities in multidimensional data. Ridgelets themselves are a multiscale transform designed to efficiently represent functions with linear discontinuities; they achieve this by applying the Radon transform to map line discontinuities to pointlike features, and then applying a one-dimensional wavelet transform along the Radon parameter. Because ridgeletlike is not a formal standard term, its precise meaning varies across sources and contexts.
Ridgeletlike approaches extend this idea by using similar principles but with different underlying transforms or dictionaries.
Applications of ridgeletlike ideas include edge detection, line feature extraction, image denoising and compression, medical imaging,