Evento CDVS: 07 - Gusmao "Localizzazione di un robot basata su analisi visuale"
0 likes368 views
Robotic localization based on visual analysis. Global and local visual descriptors are extracted from images to offer an initial estimate of a robot's position as well as the necessary precision for localization. Compression of these descriptors allows for efficient transmission in distributed robotic systems. The localization process involves extracting compressed descriptors from images taken by a robot at different times and comparing them to determine its position.
Evento CDVS: 07 - Gusmao "Localizzazione di un robot basata su analisi visuale"
- 1. Robotic Navigation ROBOTIC LOCALIZATION BASED ON VISUAL ANALYSIS Pedro Porto Buarque de Gusm達o UNINFO Milan 06/10/2014
- 2. Robotic Navigation - Localization Why?
- 3. The possibilities Autonomous Driving Precision in cm rather than m. Precise localization also for pedestrians. Intelligent Cleaning Never miss a spot
- 4. Why MPEG CDVS? Global Descriptors Offers an a first, overall estimate of robots position. Local Descriptors Offers the necessary precision for localization. Compression Allows for efficient transmission in distributed systems.
- 5. Localization Process Pavement @ t=n Pavement @ t=n- Robot pt_grey_image cdvs_extractor Workstation cdvs_compare(n,n-)CDVS Descriptors
- 7. This work is licensed under the Creative Commons Attribution- NonCommercial-NoDerivs 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/