计算光谱成像系统及光谱重建算法

Key Points

• The aim is to outline the theoretical framework and methodologies of computational spectral imaging.
• Introduced optical encoding devices to compress high-dimensional spectral data.
• Employed advanced spectral reconstruction algorithms to decode original spectral images.
• Analyzed typical optical encoding strategies like amplitude, wavelength, wavefront, and multi-aperture encoding.
• Reviewed mainstream reconstruction methods, including prior constraint-based iterative algorithms and end-to-end deep learning models.
• Demonstrated significant advantages in compactness, acquisition speed, and manufacturing costs of the imaging system.
• Highlighted the expanding applications in consumer technology like smartphones, drones, and remote sensing satellites.
• Discussed emerging trends and key challenges in the field, indicating potential impacts on smart manufacturing and agriculture.