Repository for Crop Mapping and Survey Drone project - building on previous work with drones. See meadeor-drone (Fire rescue drone) and corona-drone (test/vaccine transport drone). This project uses a bespoke stereo camera system to sense depth and build pointcloud maps of crops and land for further analysis. The outputs can help farmers understand their crops better with things like health, distribution, growth trajectory etc...
Stereo camera rig using two Coral cameras and 3d printed frame. with a custom built library for handling synced image capture and video capture. Python based library that uses gstreamer and opencv to manage image caputure and saving. Uses threading to reduce delay between left/right image pairs. Calibration, undistoriton and rectification all calculated using opencv and matlab.
Using opencv to make depth maps from stereo images using SGBM and BM. Alternatively, MiDaS and Keras ML models also used for predicting depth for when using stereo camera as a monocular system.
Converted from depth imagery. Uses RGBD data structures to make 3D meshes, voxel grids and pointcloud analysis also done using open3d.