Though many researchers have developed algorithms or techniques or methods to perform event detection, object identification and tracking, most of them have developed and tested it for a known dataset (or a publicly available dataset). Very less work has been reported on such techniques for a real-time environment. Thus in Indian context, our proposed project and research will be significant, as we are planning to develop and deploy this in an existing building, thus making it become “smarter” to cater to safety of people. In precise, this work will focus on safety and security of people and objects in a building and to connect to first responders in case of emergencies.
The proposed system includes setting up cameras and vision systems in a building, developing smart applications that run over this infrastructure and the underlying database system to record events and the videos. The major components of this project would be:
Setting up vision systems: This would be required to detect events, objects, people, identify location, and monitor the infrastructure. There are a variety of cameras and vision systems available in the market such as night vision cameras, IP cameras, thermal imaging cameras, web cameras etc. Our goal is to utilize the experience we have gained from our pervious project Indoor Information Representation and Management System (IIRMS) project funded by NRDMS, DST where some work on event detection has been done . Identifying suitable cameras and selecting the appropriate locations for day and night vision will be the highest priority.
Developing Applications: Once the infrastructure is set, we propose to develop efficient algorithms for object identification and tracking; unusual event detection and raising alarms/ warnings.
At the end of the project, we plan to have a complete study suggesting the best location for placing cameras/ vision systems for a smart building, a framework for configuring vision systems, a suite of algorithms to support object identification and tracking and specific events detection.
Key words: Event detection, image and video sequences, people tracking, vision systems
The broad objective of this project is to develop a) a framework for smart buildings that supports efficient indoor tracking algorithms for detecting events such as crowded areas, fire, smoke etc. using vision systems b) identification and tracking of moving objects over indoor spaces. This framework will complement the indoor information system developed as a part of the earlier IIRMS project funded by DST. We shall also explore integration of standards like SWE over IndoorGML and extending IndoorGML to include representation of objects and event is another important goal, which can help standardize the smart building applications. Specific objectives include the following, efficient algorithms will be developed, tested and deployed:
Specific Objective 1: Develop algorithms and build application for identifying unexpected events like locating crowded areas, fire detection, rain detection, smoke detection, electrical burst, and water burst in a given building.
Specific Objective 2: Develop algorithms and build application for identifying major assets and tracking their movement from one place to another from time to time (spatiotemporal) in this environment.
Computational thinking for Research and Education (CORE) Lab
The CORE Lab is an initiative by an Interest Group from the Computer Science and Engineering department in an attempt to transfer learning process from classrooms to practical domains. The lab offers opportunities and mentorship for students to strategically approach and solve problems the Computational Thinking way. This is the first lab of its kind in the country that sets a personal timeline for students to explore, experiment and learn by integrating theory and practice using computational thinking strategies.
For details on projects, talks, research activities, etc. please contact email@example.com
Smart Spaces Lab
With the government of India’s focus on “Digital India”, building Smart Cities is one of its major initiative. For smart cities to be built or renovated, it is essential that the existing buildings and infrastructure become smarter. In this project we aim to develop a software framework on which applications can be developed for buildings, to respond to emergencies and disaster management.