Single Pixel Imaging

Single-pixel imaging (SPI) is a computational imaging technique that reconstructs images using a single detector and structured illumination patterns, offering potential advantages in cost and miniaturization. Current research focuses on improving reconstruction quality and speed through advanced algorithms, including deep learning models like convolutional neural networks and transformers, often incorporating compressive sensing and iterative methods to handle the underdetermined nature of the problem. These advancements are enabling applications ranging from high-resolution 3D imaging and material analysis to real-time sensing and even novel applications in solving combinatorial optimization problems.