Depth Estimation from a 3D Surface Perspective for Robotic-Assisted Surgery
3D endoscopic depth estimation using 3D surface-aware constraints
Depth estimation is a representative perception task that is typically defined as an image reconstruction problem.We propose a loss function for depthestimation that integrates the surface-aware constraints, leading to a fasterand better convergence with the valid information from spatial information.In addition, camera parameters are incorporated into the training pipeline to increase the control and transparency of the depth estimation.We also integrate a specularity removal module to recover more buried imageinformation.Quantitative experimental results on endoscopic datasets and userstudies with medical professionals demonstrate the effectiveness of our method.
