Goal: I took a multi-view image of a potted plant with a handheld imaging hyperspectral camera, and I want to get a hyperspectral point cloud of the plant using Metashape reconstruction.
Problem: I first chose a clear channel as the main band, then reconstructed the sparse point cloud, then the dense point cloud. The sparse point cloud works ok, but the dense point cloud is very small and seems to be a poor depth image.
Does anyone know how I can reconstruct a hyperspectral dense point cloud with better accuracy?
Here are some screenshots:
dense point cloud:https://ibb.co/dtB6xvY
original single band image:https://ibb.co/YhrNpkh
depth image:https://ibb.co/hRsmC9C
sparse point cloud:https://ibb.co/pQGd5yv