Bo-Net is a deep learning pipeline for automated segmentation and quantification of mouse bone fluorescence microscopy images. It segments osteoblasts, osteoclasts, blood vessels, and bone matrix from confocal and multiphoton datasets using U-Net/ResNet-based models.