Segmentation of the map of depths and video image for generation of a semantic map of the environment.

The project concerns autonomous perception of a mobile robotized system. The purpose is to analyse map of depths (laser scans) and associated colour images acquired by the robot during movement in indoor environment.

The project concerns autonomous perception of a mobile robotized system. The purpose is to analyse map of depths (laser scans) and associated colour images acquired by the robot during movement in indoor environment. It is assumed that results of analysis of such data will take the form of a 3D semantic map of the environment. Data will be captured for the following two scenarios of robot operation:

  1. Bypassing obstacles during travel through a corridor to the target location;
  2. Picking up a selected object from among more objects lying on a table or floor.

The project distinguishes the following research topics:  

  • Generation of a 3D map by merging data from the map of depths and from the video image;
  • Segmentation of the 3D map and classification of selected objects.

Project implementation timeframe: 01/07/2012 – 15/12/2012

Coordinator Industrial Research Institute for Automation and Measurements PIAP