Abstract
In advanced single-pixel imaging systems that use spatial light modulators (SLMs), the scene needs to be illuminated by numerous different patterns that are coded into the SLM to gather information about the object. The appropriate choice of pattern set and reconstruction algorithm are essential for rapid reconstruction of a high-quality image. For single-pixel quantum ghost imaging (QGI) systems, we introduce an adaptive pattern generation method based on the generalized Ising model. The 2D Ising model describes an array of spins arranged in a lattice, where each spin in the lattice can interact with its adjacent spins and also can be influenced by an external magnetic field. By modeling spin values (up or down) as pixel states of the SLM (on or off) and using the Ising model, we can generate binary patterns that closely resemble the object's shape. Imaging using patterns generated by our proposed method can reduce the number of measurements required for high-quality image reconstruction.