This study presents a near-infrared (NIR) hyperspectral imaging system that uses single-pixel detection and self-assembled colloidal quantum dots (CQDs) for spectral and spatial information encoding. The system replaces traditional InGaAs focal plane arrays (FPAs), which are expensive and bulky, with a more cost-effective and compact solution. The CQDs are used to modulate the spectral information, while a digital micromirror device (DMD) encodes the spatial information. The single-pixel detector then reconstructs both the spectral and spatial data using compressed sensing algorithms. The system achieves a detection window of about 600 nm and an average spectral resolution of 8.6 nm with a pixel resolution of 128 × 128. The results show that the system provides a high signal-to-noise ratio (SNR) for both spectral and image reconstruction, outperforming conventional FPA-based systems. The spectral and spatial data align well with reference instruments, validating the effectiveness of the approach. The system offers a promising solution for affordable and accessible NIR hyperspectral imaging, expanding the range of potential applications. The study also demonstrates the system's ability to capture hyperspectral images of emissive, transmissive, and reflective objects, including real and plastic grasses, and fresh strawberries. The results show that the system can accurately identify and characterize materials based on their spectral signatures, with high resolution and accuracy. The system has potential applications in various fields, including pharmaceuticals, food processing, and cultural heritage preservation. The study highlights the advantages of using CQD filters and single-pixel detection for NIR hyperspectral imaging, offering a more efficient and cost-effective solution compared to traditional methods.This study presents a near-infrared (NIR) hyperspectral imaging system that uses single-pixel detection and self-assembled colloidal quantum dots (CQDs) for spectral and spatial information encoding. The system replaces traditional InGaAs focal plane arrays (FPAs), which are expensive and bulky, with a more cost-effective and compact solution. The CQDs are used to modulate the spectral information, while a digital micromirror device (DMD) encodes the spatial information. The single-pixel detector then reconstructs both the spectral and spatial data using compressed sensing algorithms. The system achieves a detection window of about 600 nm and an average spectral resolution of 8.6 nm with a pixel resolution of 128 × 128. The results show that the system provides a high signal-to-noise ratio (SNR) for both spectral and image reconstruction, outperforming conventional FPA-based systems. The spectral and spatial data align well with reference instruments, validating the effectiveness of the approach. The system offers a promising solution for affordable and accessible NIR hyperspectral imaging, expanding the range of potential applications. The study also demonstrates the system's ability to capture hyperspectral images of emissive, transmissive, and reflective objects, including real and plastic grasses, and fresh strawberries. The results show that the system can accurately identify and characterize materials based on their spectral signatures, with high resolution and accuracy. The system has potential applications in various fields, including pharmaceuticals, food processing, and cultural heritage preservation. The study highlights the advantages of using CQD filters and single-pixel detection for NIR hyperspectral imaging, offering a more efficient and cost-effective solution compared to traditional methods.