Monte-Carlo Based Efficient Image Reconstruction in Coherent Imaging With Speckle Noise

Speckle is a multiplicative form of noise that poses a significant challenge in coherent imaging systems. Although maximum likelihood estimation (MLE) can be used to mitigate speckle noise, when dealing with realistic imaging systems (finite apertures) the associated high-dimensional matrix inversions make the computational costs of MLE prohibitively expensive at high resolutions. This work introduces a novel approach to perform MLE without explicit matrix inversions, which 1) significantly speeds up reconstruction; 2) enables high-resolution image recovery; and 3) maintains robust performance under high additive noise and realistic aperture models.