RAFI & RAFI- II - Autonomous Area Mapping and Multi-tasking Robot
Description:
- This robot is equipped with two webcams in the head for 3D vision.
- It has two hands with 5 degrees of freedom for grabbing objects and it can move avoiding obstacles analyzing sonar-data.
- The head has 3 degrees of freedom. It has the power of object-recognition as it is equipped with proper object-matching algorithm.
- It can sense the presence of darkness and it can use its own light-source if needed for mapping.
Computer-vision algorithms are implemented in MATLAB.
This robot can perform following tasks:
- Object grabbing
- Object recognition- Automatic object matching in real time using MSER technique to match the regional properties at first and then validating them using SIFT algorithm. It can match the objects irrespective of scaling and rotational variation.
- 3-D mapping of the object’s surrounding as well as Anaglyph images creation by Image rectification- Two images, taken with parallel cameras in real time, are processed to create anaglyph and rectified images using stereo image rectification algorithm. Then binocular stereo vision algorithm is used to generate 3D map of the surrounding environment.
The code is shared in my github page in this link: https://github.com/mdshafayat/RAFI