• Sensor Fusion-Implemented Velodyne Lidar -16 and Monocular Camera sensor fusion using extrinsic calibration to find rigid body transform
between the sensors of ARAI’s Swayam Go vehicle.OpenCV’s PnP RANSAC method was used to find the rotational and translations of Lidar and
Camera. Appropriate transformation were applied to get data fusion.
• Displaying fused Lidar points on camera image using OpenCV’s geometric projection on pinhole camera model.
• Implemented YOLO object detection to draw bounding box in image and project the bounding box detected in Image and transform coordinates of image in Lidar . 2D boxing in 3D world.
• Using Turtlebot for SLAM and Navigation Algorithms
• Worked on Data acquisition of Camera,Lidar,GPS from ARAI’s Autonomous Vehicle ’Swayam Go’.
• Implemented Obstacle detection on Point Cloud Data collected by the vehicle to detect obstacles in the path.This will include RANSAC segmentation of road,plane and obstacles and clustering algorithm in PCL.
• Developed Custom Object Detector using YOLO framework for detecting vehicles specific like Traffic Signs,Rickshaws,bikes and Tempos.
Tools Used- ROS,PCL,OpenCV,Numpy,C++