Block-based single-pixel imaging by means of the Talbot effect

Block-based single-pixel imaging (SPI) systems use multiple detectors of a focal plane array to create highly parallel extensions of a single-pixel camera. This Letter presents what we believe to be a new approach for parallel single-pixel imaging that leverages the Talbot effect to scan an object with a short sequence of periodic light patterns encoded on a digital micromirror device (DMD). The unit cell of each periodic pattern corresponds to a Walsh-Hadamard (WH) matrix of low dimension. To separately collect the light from each unit cell, we implement an array of bucket detectors using multiple pixels of a conventional camera. The object image is then reconstructed by applying parallel single-pixel imaging algorithms. Our system improves spatial resolution and the frame rate while eliminating the need for optical elements between the DMD and the object. Simulations and experiments with two optical setups are presented and analyzed.