NSITonline

Dynamic scene management is a process by which changes in a dynamic scene, such as a video, can be handled using data structures. Its everyday applications are widespread - right from streaming online videos to rendering good animation graphics.

Tracking, monitoring and management of moving objects become increasingly important in modern geospatial application. An important example of such applications is monitoring and management of a dynamic scene captured by a network of videos or sensors for ensuring for example the security of the people during a public event, or for leading emergency or rescue teams in real environment.

At present the most important things on scene rendering in computer graphics are fidelity and efficiency. In order to achieve these goals, a considerable number of applications require a data structure like a tree to manage the entire scene. Searching the scene means to traversal the tree and moving objects in the scene means to recreate the tree, which are very costly but unavoidable process.

This project basically seeks to make this process faster, while maintaining the quality of the output graphics. Possible applications for this algorithm would be in fast graphics, such as in arcade and online gaming. 

Finally, it discuss other potentials and limitations of the proposed model and suggest some new research avenues regarding the improvement of the proposed model.

This project has been shaped by Vaishnavi Sundararajan and Saurav Malviya of Instrumentation and Control Department of NSIT, graduating in the year 2010. This has been achieved under the guidance of Prof. Anand Gupta, an esteemed Graphics lecturer in the Computer Science Division.

This project, as they both agree, is the longest running and most interesting project of their undergraduate studies so far. This project involves the management of dynamic scenes in three dimensions using tree data structures called octrees. Octrees are most often used to partition a three dimensional space by recursively subdividing it into eight octants.

This project aims at improving the existing resource-hungry methods of dynamic scene management in three dimensions. The previously existing method involved resizing of the nodes of an octree, while this project completely avoided or minimised it, thus achieving greater efficiency. The algorithm developed proved to be much faster than the existing ones (for certain scene configurations, as much as 25% faster).

They even developed a research paper on the same, which was published at the IEEE International Multitopic  Conference 2008.

After having published this paper, they came up with a more general method which could handle a greater variety of scenes than their previous algorithm, while maintaining the same time complexity. This method is the focus of the transaction which is currently under review at the IEEE Image Processing Society.

As far as real world applications are concerned, this method can be used for any situation involving changes. Possible applications of this project are face isolation and face following.

Vaishanvi S. can be contacted at This e-mail address is being protected from spambots. You need JavaScript enabled to view it .