A Framework for Dynamic Image Sampling Based on Supervised Learning

Sparse sampling schemes can broadly be classified into two main categories: static sampling where the sampling pattern is predetermined, and dynamic sampling where each new measurement location is selected based on information obtained from previous measurements. Dynamic sampling methods are particularly appropriate for pointwise imaging methods, in which pixels are measured sequentially in arbitrary order. Examples of pointwise imaging schemes include certain implementations of atomic force microscopy, electron back scatter diffraction, and synchrotron X-ray imaging. In these pointwise imaging applications, dynamic sparse sampling methods have the potential to dramatically reduce the number of measurements required to achieve a desired level of fidelity. However, the existing dynamic sampling methods tend to be computationally expensive and are, therefore, too slow for many practical applications. In this paper, we present a framework for dynamic sampling based on machine learning techniques, which we call a supervised learning approach for dynamic sampling (SLADS). In each step of SLADS, the objective is to find the pixel that maximizes the expected reduction in distortion (ERD) given previous measurements. SLADS is fast because we use a simple regression function to compute the ERD, and it is accurate because the regression function is trained using datasets that are representative of the specific application. In addition, we introduce an approximate method to terminate dynamic sampling at a desired level of distortion. We then extend our algorithm to incorporate multiple measurements at each step, which we call groupwise SLADS. Finally, we present results on computationally generated synthetic data and experimentally collected data to demonstrate a dramatic improvement over state-of-the-art static sampling methods