top of page (2).gif

Internship : Perception for Robotics


I was selected for the 6  month Internship at the Automotive Association of India .

I developed a perception  pipeline for ARAI’s Autonomous Vehicle ‘Swayam GO’ .

  1. Setup and Calibrating LiDAR- Camera

  • Setting  up LiDAR

  • Setting up Camera

  • Intrinsic Calibration of Camera

  • Data acquisition from LiDAR and Camera


Sensor Fusion

  • Transformation Matrix between LiDAR and Camera

  • Acquiring co-ordinates of board corners from Camera.

  • Acquiring co-ordinates of board corners from LiDAR.

  • Finding the correspondence between LiDAR and Camera.

  • Projecting the 3D LiDAR points to 2D Image.

  • Projecting 2D Image in 3D LiDAR data.


3D Obstacle Detection in LiDAR.

  • Crop Box function

  • Plane Segmentation using RANSAC.

  • Euclidean Clustering

  • Min-Max Boxing


Custom Object Detection  .

  • Data Preprocessing

  • Data collection for Indian Dataset.

  • Architecture Selection.

  • Hyper parameters tuning and Training the Model.

  •  Testing and Improving Models.


Checkout the implementation on GitHub

bottom of page