A real-time ball trajectory follower using Robot Operating System
0 likes118 views
The document presents a study on a real-time ball trajectory follower using the Robot Operating System (ROS) aimed at enhancing sports event coverage with Unmanned Aerial Vehicles (UAVs). It outlines the methodology, system description, and experimental results demonstrating the effectiveness of image processing techniques like color segmentation and object tracking. The findings indicate high detection accuracy and processing speed for tracking the ball during sports events.
A real-time ball trajectory follower using Robot Operating System
- 1. A real-time ball trajectory follower using Robot Operating System Himanshi Yadav National Institute of Technology, Tiruchirappalli, India Siddharth Srivastava, Prerana Mukherjee and Brejesh Lall Department of Electrical Engineering, Indian Institute of Technology, Delhi, India
- 2. Outline • Introduction • Brief Methodology • System Description • Proposed Approach • Experimental Results • Conclusion • Questions
- 3. References • Superhuman Sports Society aims to bring Harry Potter’s Quidditch, Dragon Ball to life by Yuzuha Oka http://www.japantimes.co.jp/news/2015/06/03/business/te ch/superhuman-sports-society-aims-bring-harry-potters- quidditch-dragon-ball-life/#.VnUA_vl97IV
- 4. Introduction http://img.thesun.co.uk/aidemitlum/archive/01520/stadium _1520387a.jpg
- 7. Use of UAVs (Unmanned Aerial Vehicles) for sports events http://a2.fssta.com/content/dam/fsdigital/fscom/MOTORS/i mages/2015/03/16/drone1.vadapt.620.high.68.jpg
- 8. Why UAVs? • Inexpensive • Autonomous flight: lesser human involvement • Closer and real time view: allows better tracking of the ball and players
- 10. System Description •Mechanical System •Software System Robot Operating System(ROS) interface Image Processing Control System
- 11. Proposed Approach • Image retrieval • Image processing Colour Segmentation Morphological Operations Object Recognition using image moments Object Tracking • Control System ://opencv-python-tutroals.readthedocs.org/en/latest/py_tutorials/py_imgproc/py_contours/py_contour_features/py_contour_features.html#contour-features
- 12. Object Tracking Error in position= Centre of object detected - Image view centre
- 13. Control System
- 14. Experimental Results • Ball Detection
- 15. Experimental Results • Ball Detection
- 16. Experimental Results • Detection accuracy with change in ball velocity
- 17. Experimental Results • Processing Speed Operation Time taken(in milliseconds) Image acquisition 1 Colour segmentation 1.5 Ball localisation 1.2 Total 3.7
- 19. Questions?