Matching Pursuit Decomposition (MPD) is a powerful and effective iterative algorithm for signal decomposition and feature extraction. MPD decomposes any signal into linear combinations of its dictionary elements or “atoms”. However, a sufficiently large dictionary is required for an accurate reconstruction; this in return increases the computational inefficiency of the algorithm, thus limiting its applicability and level of adoption.
MPD++ was developed to improve the scalability and performance of the classical MPD algorithm. The proposed modifications include correlation thresholding, multiple atom extraction and course-fine grids technique which are capable of being stacked and have cumulative effects on the reduction of the run time complexity of MPD algorithm.