This study presents a novel approach to near-infrared (NIR) hyperspectral imaging using single-pixel detection and self-assembled colloidal quantum dots (CQDs). The high cost of InGaAs focal plane arrays (FPAs) has hindered the widespread adoption of NIR hyperspectral imaging. To address this, the researchers employ a single-pixel detector in conjunction with CQD filters and a digital micromirror device (DMD) for spectral and spatial information multiplexing. The experimental results demonstrate successful NIR hyperspectral imaging with a detection window of about 600 nm, an average spectral resolution of 8.6 nm, and a pixel resolution of 128 × 128. The system shows superior noise resilience compared to conventional FPA, with a significant improvement in signal-to-noise ratio (SNR) for both spectral and image reconstruction. The spectral reconstruction capabilities of the system are enhanced by the unique absorption characteristics of CQDs, which offer better spectral resolution and noise tolerance compared to edge-pass and band-pass filters. The system's effectiveness is validated through comparisons with commercially available spectrometers and applications in identifying emissive, transmissive, and reflective objects, such as LEDs, transparent solutions, and fresh strawberries. The study highlights the potential of this approach for affordable and accessible NIR hyperspectral imaging technologies, expanding their range of potential applications.This study presents a novel approach to near-infrared (NIR) hyperspectral imaging using single-pixel detection and self-assembled colloidal quantum dots (CQDs). The high cost of InGaAs focal plane arrays (FPAs) has hindered the widespread adoption of NIR hyperspectral imaging. To address this, the researchers employ a single-pixel detector in conjunction with CQD filters and a digital micromirror device (DMD) for spectral and spatial information multiplexing. The experimental results demonstrate successful NIR hyperspectral imaging with a detection window of about 600 nm, an average spectral resolution of 8.6 nm, and a pixel resolution of 128 × 128. The system shows superior noise resilience compared to conventional FPA, with a significant improvement in signal-to-noise ratio (SNR) for both spectral and image reconstruction. The spectral reconstruction capabilities of the system are enhanced by the unique absorption characteristics of CQDs, which offer better spectral resolution and noise tolerance compared to edge-pass and band-pass filters. The system's effectiveness is validated through comparisons with commercially available spectrometers and applications in identifying emissive, transmissive, and reflective objects, such as LEDs, transparent solutions, and fresh strawberries. The study highlights the potential of this approach for affordable and accessible NIR hyperspectral imaging technologies, expanding their range of potential applications.