My research focuses on developing a multispectral camera system for labeled and label-free fluorescence cancer imaging. Inspired by the unique structure of the mantis shrimp’s eye, this camera can simultaneously capture images across the UV, visible, and NIR spectrums. The system has significant implications for advancing image-guided surgery and intraoperative pathology. Based on the bioinspired camera, I also developed a lensless microscopy within the UV-Visible-NIR spectrum. By replacing expensive optical lenses with a holographic diffuser, the lensless microscopy achieves even higher resolution and enables three-dimensional imaging with a single shot. To optimize the bioinspired camera’s image demosaicing, I designed and trained a 20-layer convolutional network with residual learning, significantly enhancing reconstructed image quality compared to traditional methods.