Image Tagging for Autonomous Car

Recent increase in the adoption of devices for capturing digital media and mass storage systems have led to an explosive number of images and videos. As a part of our image tagging and machine learning module we have been spending a lot of time in the development of appropriate techniques for robust lane detection, object classification, tracking, and representation of task relevant objects.


List of image processing we have undertaken under image tagging and machine learning module,

  • Tagging vehicles on road by packing them in a tight rectangular boundary box
  • Lane marking: marking lanes on the road, and distinguishing the types of lanes and colors of the lanes
  • Surface tagging: tagging visible road area by excluding vehicles
  • Segmentation: tagging complete number of objects seen in the image, like road, lanes, vehicles, traffic signs, nature, road barriers, trees and others which are seen on the road side
  • Traffic signs: tagging traffic sign boards in the shape which is present on the road, road side or above and also transcribing the text mentioned on the sign boards
  • Vehicle tracking: annotating vehicles in tight rectangular box and tracking, also unique identification of the cars on the road/drive way