Difference between revisions of "Dense 3D Reconstruction"

From AIRWiki
Jump to: navigation, search
 
Line 11: Line 11:
 
| image= }}
 
| image= }}
  
3D Reconstruction is a well investigated research topic in the Computer Vision and Robotics areas. Recent publications prove the ability of current state of the art systems to address the problem of reconstructing the 3D shape of objects in very different contexts and scale. Yet, there is still a large margin of possible improvement in accuracy, performance and adaptability to the different environments. The goal of this project is to investigate novel state of the art approaches, and their implementation on different hardware platforms, targeting graphics processing units in particular.
+
3D Reconstruction is a well investigated research topic in the Computer Vision and Robotics areas. Recent works prove the ability of state of the art systems to address the problem of reconstructing the 3D shape of objects in very different contexts and scale. Yet, there is still a large margin of possible improvements in accuracy, performance and adaptability to the different environments. The goal of this project is to investigate novel state of the art approaches, and their implementation on different hardware platforms, targeting graphics processing units in particular.

Latest revision as of 23:48, 5 June 2016

Dense 3D Reconstruction
Short Description: The goal of the project is researching state of the art techniques in the field of 3D Reconstruction
Coordinator: MatteoMatteucci (matteo.matteucci@polimi.it)
Tutor: AndreaRomanoni (andrea.romanoni@polimi.it)
Collaborator:
Students: AndreaBignoli (andrea.bignoli@mail.polimi.it)
Research Area: Computer Vision and Image Analysis
Research Topic: 3D Reconstruction
Start: 2016/06/05
Status: Active

3D Reconstruction is a well investigated research topic in the Computer Vision and Robotics areas. Recent works prove the ability of state of the art systems to address the problem of reconstructing the 3D shape of objects in very different contexts and scale. Yet, there is still a large margin of possible improvements in accuracy, performance and adaptability to the different environments. The goal of this project is to investigate novel state of the art approaches, and their implementation on different hardware platforms, targeting graphics processing units in particular.