Computational imaging system with outdoor natural light based on Hadamard Transform and Compressive Sensing

This paper describes a single-pixel-detection computational imaging approach using outdoor natural light. We produced an imaging prototype based on digital micro mirror (DMD). The prototype is equivalent to an optical computing system. Measurement of the input scene through band pass filter is equivalent to projective measurement in the transform domain, and hence can be treated with the Hadamard Transform or the Compressive Sensing frameworks recently developed by a number of groups. Computational imaging system offers a versatile approach to detect images' signal which can improve size, weight and easily controllable compared with traditional imaging approaches. Finally, we conducted some indoor and outdoor experiments for our imaging prototype, and also analyzed the differences between Hadamard Transform imaging and compressive sensing imaging.