An infrastructure for interactive environments
Abstract
Description
A thesis Submitted in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE in COMPUTER SCIENCE from Texas A&M University-Corpus Christi in Corpus Christi, Texas.
We present an infrastructure for interactive environments. The result of our implementation is a fast and reliable system used to recognize and track people in indoor surroundings. The system is aware of the location and identity of people, then can interact with the human or monitor their activity. Our system is combined with three main components, which are body tracking, face recognition, and the controller. The controller is responsible for collecting and matching the information from body tracking and face recognition, and give command signals to interaction and monitoring modules. To ensure the system works effectively, it is required to meet two conditions. First, the tracking and the recognition modules need to be fast and accurate. Second, there is a method to match faces with bodies efficiently. Our work provides solutions for those demands. The system deploys RGB cameras and Kinect depth sensors to obtain human information as the input. Overall, the people recognition-tracking of the system works at around 10 fps, and the system can make some basic interactions.
Computing Sciences
College of Science and Engineering
We present an infrastructure for interactive environments. The result of our implementation is a fast and reliable system used to recognize and track people in indoor surroundings. The system is aware of the location and identity of people, then can interact with the human or monitor their activity. Our system is combined with three main components, which are body tracking, face recognition, and the controller. The controller is responsible for collecting and matching the information from body tracking and face recognition, and give command signals to interaction and monitoring modules. To ensure the system works effectively, it is required to meet two conditions. First, the tracking and the recognition modules need to be fast and accurate. Second, there is a method to match faces with bodies efficiently. Our work provides solutions for those demands. The system deploys RGB cameras and Kinect depth sensors to obtain human information as the input. Overall, the people recognition-tracking of the system works at around 10 fps, and the system can make some basic interactions.
Computing Sciences
College of Science and Engineering