Enhancing Compressive Single-Pixel Imaging With Zig-Zag-Ordered Walsh-Hadamard Light Modulation

Single-pixel imaging (SPI) is empowered to achieve high-quality image reconstruction by utilizing only single-element detectors via compressive sampling in a high-efficiency, low-cost and low-complexity manner. This letter proposes an easy-to-implement demonstration of SPI with zig-zag-ordered Walsh-Hadamard spatial patterns to eliminate artifacts and enhance the image quality of the reconstructed scene. Dual-channel measurements are taken and integrated to improve the signal-to-noise ratio (SNR) and enhance the fidelity of the images. A calibration procedure of the pixels-to-detector ray propagation is experimentally validated to reveal the true reflectance of the scene. Compressive Sensing is incorporated to reconstruct high-fidelity images with sub-Nyquist compression ratios (CRs). Experimental results of the proposed approach demonstrate above 90% correlation with CRs down to 30% in a completely dark environment. This work provides insights for efficient high-fidelity SPI in low-light or low-signal-to-noise-ratio (low-SNR) conditions.