Projects introducing the concept of sensor fusion, from learning how to process three types of car sensor types:
- Lidar: Process point clouds to segment, clusterize and estimate distances. Final project involved process a short video of point cloud and clusterize all objects in the scene.
- Camera: Tutorials on point detection, description and matching using OpenCV to track individual objects through and video scene and estimate their distance using the image's size variation.
- Radar: How radar perceives the world and distance, angle and speed estimations using Doppler Response and Constant false alarm rate - CFAR.
and to fuse all of their informations to produce final more reliable estimates with
- Unscented Kalman Filter: What are kalman filters, how to process the two steps of predict and estimate, what are extended kalman filters and their linearization process and, finally, what are unscented kalman filters.
