Computer vision based self-recharging robot

Published on 20 May 2021. Written by Athulya Menon

Computer vision is one of the leading technologies which is undergoing a lot of revision, and there is a lot of research happening in this field as well. It essentially gives computers, devices and other machines the ability to see and perceive images. From a pure engineering point of view, technology such as

Read more..

computer vision helps in the creation of recognition systems that will one day mimic human vision. The field has a lot of real-world applications, and is hence, one of the most important fields of study in the present scenario. Here’s a look at a recharge platform for autonomous robots, which will help them charge their batteries.

Evolution of Computer Vision

Computer vision has its origins in the 1960s, where it started out as a thesis idea by a few

MIT students. Their initial idea was to extract 3D geometrical properties from 2D blocks, giving rise to the concept of computer vision. In the 1970s, scientists began using these concepts in real-world applications and started to analyse images of the world around us. At this time, most of the research focused on edge detection and segmentation. When the 1980s came along, computer vision shifted towards a more geometry-centric approach and started utilising mathematical concepts such as regression. Finally, in the 1990s, face recognition systems became a huge part of such systems, and Computer vision became a staple for most security systems. Fast forward to today’s times, computer vision is an integral part of our day-to-day lives. It has found large-scale applications in the medical field, manufacturing processes, security systems and also automation.

Project Description

An upcoming field that has tremendous scope is that of autonomous vehicles, and these include all kinds, such as aerial, terrestrial and aquatic vehicles. Such systems make use of computer vision technology to navigate and also to detect obstacles and targets. In this project, we will be using computer vision to help such an unmanned vehicle find a charging platform and dock itself to the charging, allowing itself to recharge its batteries. This will allow such bots to last for longer periods of time, and hence cover larger sections of land.

Concepts Used

1. Basics of mechanics
2. Robotics and Automation
3. Python Programming
4. Computer Vision Fundamentals
5. Image Processing
6. Data Segmentation

Hardware Required

• Robot of any kind
• External power source
• Charging platform with the socket or wireless charging mechanism
• Raspberry Pi cameras
• Wires and connectors
• Auxiliary power sources
• Microprocessors or Arduinos to program the robot
• Desktop to code; installed Python 2 and higher
• Supporting Python libraries installed
• OpenCV extension installed

Project Implementation

1. The robot utilizes a Raspberry Pi chip or Arduino to achieve wireless control.
2. This controller is attached to the various servo motors that allow the robot to move.
3. The controller is also connected to the cameras on the robot so that the robot can send visual data to the analyzer or desktop.
4. Next, we must set up the charging platform some distance away from the robot.
5. The charging platform must work either via a charging port or through wireless charging.
6. If it makes use of wireless charging, the robot will just have to walk into the platform to start getting charged, while if it via cables, the robot will have to plug the port into its charging dock.
7. Next, we will make use of a basic object detection algorithm using the color tracking method, to sense the presence of the charging dock.
8. The navigation system we use is fairly simple, utilising guideways for the robot to move along.
9. To help the robot navigate towards the targeted object, lines are drawn in the visual feed to act as guideways.
10. Hence, when the target appears, either it is to the left of the lines, towards the right of the lines or between both the lines.
11. The robot can then use this data to center itself and work its way towards the target with ease.
12. The computer vision program will constantly track the position of the target and ask the robot to turn as per the requirement.
13. Distance to and from the object will be tracked using the Y coordinate of the lines.
14. This way, the robot will slowly, yet surely make its way towards the charging platform to recharge itself.
15. Also, the provision will be provided to ensure that if the target is not within the scope of view of the camera, the program will ask the robot to turn and move around till it spots the target.
16. To create the vision program, students can make use of SURF, which is Speeded Up Robust Features, a popular vision target detection algorithm.

Kit required to develop Computer vision based self-recharging robot:

• No kit required

Technologies you will learn by working on Computer vision based self-recharging robot:

• Computer vision
• Image Processing
• Python

Any Questions?