Compressive Sensing for Non-stationary Music Signal & Image Signal

Compressive Sensing (CS) is the key solution or method to reconstruct the signal with very few number of measurements as compared to conventional methods. According to the conventional methods or we can say Shannon-Nyquist sampling theory we require twice the signal bandwidth for proper reconstruction of signal. The basic problem to store a large amount of data with the conventional method. To achieve this we require the measurement matrix which should be a stable measurement matrix and the basis matrix. The measurement and the basis matrix should satisfy two properties which are RIP and iid. The measurement matrix which is generally a Random Matrix is optimized to achieve smaller mutual coherence. Here are various reconstruction algorithms exist which are used for the proper reconstruction of the signal after the compressions.