Shearlets

ShearLab (Shearlets) is an open-source mathematical software platform developed at Technische Universität Berlin that implements digital shearlet transforms for processing 2D and 3D signals. Shearlets are a directional representation system extending the wavelet framework, designed to efficiently encode anisotropic features such as edges and curvilinear singularities in images and videos. The platform provides MATLAB, Python (pyShearLab), Julia (Shearlab.jl), and TensorFlow (tfShearlab) implementations, with GPU acceleration via CUDA, supporting applications like image denoising, inpainting, geometric separation, and edge analysis.