Project funded in 2010 by the National Program of Private-Public Cooperation, subprogram INNPACTO the Spanish Ministery of Economy and Competitivity With Project number: IPT-430000-2010-52
The project aims at developing a robust and no-invasive tracking system being able to process, visualize and reconstruct the 3D information about people movements (including trajectories and the positioning of relevant body parts) in a limited, indoor environment.
In order to improve the accessibility for the elderly to their own households and make easier to them the use of new technologies and house devices, it is necessary to study their habits at their homes. For this reason, the goal of the project is to develop a tracking system to register people movements in a limited, indoor space. These data will be stored digitally and will contain relevant parameters such as the 3D coordinates of people positions, the positioning of different body parts (e.g. head, shoulders, hips and limbs), as well as other information helping recognize the detected person. To get a tracking system the less invasive as possible, it is required that people are not wearing any additional marker or devices. This way, people will be able to carry out their daily activities in a natural (normal) manner.
The role played by CeDInt was the development of the tracking system through a network of Kinects™, with the aim of robustly collating spatiotemporal information about people and allow a realistic 3D representation of their movements.
The main results obtained by CeDInt are described as follows:
- Markerless tracking system for large environments using multiple KinectsTM.
Figure 1. The scheme detailing the tracking system with several KinectsTM developed for the project.
Figure 2. Tracking of multiple users: each person is characterized by a unique set of points and joints and its 3D combination provide information about his/her movements
Figure 3. A single user tracked through two different KinectsTM. This double point of view provides more 3D information with respect to a single image. At the same time, this configuration extends the trackable area, so it would be easier to follow people under different environment configurations.
- Web Application to visualize and analyze the data obtained by the tracking system:
Figure 4. Interface of the 3D web viewer where it is possible to visualize and use the information obtained by the tracking system
Video 1. A video describing the results obtained during the project.
C. A. Ortiz, B. Ríos, D. Garrido, C. Lastres, and I. Galloso, "Tracking multiple humans in large environments: proposal for an indoor markerless tracking system", in Joint Virtual Reality Conference of ICAT-EGVE-Euro VR, Current and Future Perspectives of Virtual Reality, Augmented Reality and Mixed Reality: Industrial and Poster Track, 2012, pp. 97-100.