Multispectral array detectors based on computational imaging techniques

Aiming at the problem that traditional multispectral array detectors rely on complex optical components such as filters, which leads to large system volume, high maintenance cost and limited imaging quality, this study proposes a multispectral array imaging system based on computational imaging technology. By integrating a multi-channel single-pixel detector array, combining the DMD high-speed modulation technique and differential correlation algorithm, a spectral segment parallel acquisition mechanism is constructed to realize the synchronous correlation imaging of multispectral dimensions, which avoids the limitations of the traditional hardware stacking architecture. The designed low-noise pre-amplification and filtering circuits effectively improve the signal-to-noise ratio of weak photocurrent signals, and combined with the suppression of statistical noise by the differential correlation algorithm, the imaging quality of highly reflective or transparent objects is significantly improved. The experimental results show that the system achieves four-channel simultaneous imaging in the spectral range of 400 nm~1700 nm with a 20% improvement in the signal-to-noise ratio compared with a single detector, which provides a new paradigm of lightweight and high efficiency for acquiring multidimensional information of complex scenes in the fields of environmental monitoring, precision agriculture, and medical diagnosis.