Block scanning single-pixel imaging at the single photon level in a near-infrared band

Single-pixel imaging (SPI) is highly effective for weak light detection. However, SPI often requires extensive photon accumulation cycles to enhance the signal-to-noise ratio (SNR) of images. This process can lead to pile-up effects when multiple photons hit the detector simultaneously, causing signal distortion and degradation of imaging quality. To address these challenges, we propose a blocky basis scanning method that combines the robustness of point-scanning imaging with the efficiency of SPI. This approach significantly reduces the required dynamic range for photon counting compared to SPI using Hadamard patterns. Experimental results demonstrate that the proposed strategy decreases photon usage by 25 times compared to Hadamard patterns. Additionally, the contrast-to-noise ratio (CNR) of the retrieved image is increased by 1.1 times, achieving imaging with just 1.34 photons per pixel. The proposed method improves information acquisition speed and SNR in SPI while enabling three-dimensional object reconstruction, making it promising for mapping applications.