The so-called Visual SLAM techniques have achieved a huge breakthrough in the last years. This algorithms use a single camera to simultaneously locate a robot in 3D with a great accuracy and create a map of its surroundings.
Although many other sensors can be employed with SLAM algorithms, cameras have certain advantages over other sensors:
- They are relatively inexpensive.
- They can provide enormous amounts of information.
- They can be found in many devices, such as cell phones or most of robots.
Visual SLAM algorithms have a growing number of real applications, serving as a starting point to perform more complex tasks:
- Accurately locate a robot inside an environment to adapt its behavior
- Create 3D maps that will be used to other tasks: navigation, environment exploration, etc.
- Perform augmented reality in real-time