• Process raw lidar data with filtering, segmentation, and clustering to detect other vehicles on the road.

• Fuse camera images together with lidar point cloud data. You'll extract object features, classify objects, and project the camera image into three dimensions to fuse with lidar data.

• Analyze radar signatures to detect and track objects. Calculate velocity and orientation by correcting for radial velocity distortions, noise, and occlusions.

• Fuse data from multiple sources using Kalman filters, and build extended and unscented Kalman filters for tracking nonlinear movement.

• Implemented RANSAC algorithm to separated ground plane from obstacles
• Euclidean clustering algorithm is used to identify obstacles. KD tree implemented from scratch is used to speed up searching point cloud

Lidar Obstcle Detection

Tracking a bicyclist riding in front of the car

Tracking a bicyclist riding in front of the car

Various combination of keypoint detectors, descriptors and matching schemes are explored.

2D_Features

• Various combination of keypoint detectors, descriptors and matching schemes are explored.
• Object detection using the pre-trained YOLO deep-learning framework
• Methods to track objects by matching keypoints and bounding boxes across successive images
• Associating regions in a camera image with lidar points in 3D space

TTC Calculation