search

gpcv-liujin / Deep3D_Aerial

This repository provides the official implementation of the paper *Deep learning based multi-view stereo matching and 3D scene reconstruction from oblique aerial images*.
GNU General Public License v3.0
7 stars 1 forks source link

fuse depth_map #1

Open wln19 opened 2 months ago

wln19 commented 2 months ago

Thank you for your excellent work. When merging point clouds, the ply visualization colors cannot be loaded correctly. Do I need to set additional parameters? scene

gpcv-liujin commented 2 months ago

Thank you for your excellent work. When merging point clouds, the ply visualization colors cannot be loaded correctly. Do I need to set additional parameters? scene

The error you encountered while saving the PLY file is due to using the wrong data type. The color data should be saved as uint8.

This is mostly attributed to using a new version of the numpy library. The expression color = (color * 255).astype(int) needs to be updated. Replace np.int with np.int32 to ensure compatibility with the new numpy version.

wln19 commented 2 months ago

Thank you for your answer. We also encountered some problems in generating Mesh. Which version of openmvs did you install? After we installed it, we kept getting errors when running the code. 捕获

gpcv-liujin commented 1 month ago

Thank you for your answer. We also encountered some problems in generating Mesh. Which version of openmvs did you install? After we installed it, we kept getting errors when running the code. 捕获

We used OpenMVS1.1.1. You can kindly ingore this error if the mesh result can be output .