Video analysis for assessing hatchling health condition
The goal of this project is to device automated systems to assess sea turtle hatchlings' health condition through video analysis. This research is being carried out in collaboration with Juara Turtle Project, Pulau Tioman, Malaysia.
The goal of this project is to device an automated species identification system from photos of sea turtle carapaces. The resulting solution can be deployed wide range of computing platforms starting from portable devices like smart phones to high end workstations. This problem would be classified in the realms of Computer vision or Image processing. This is an ongoing project. At the time of writing this, the system can detect Green and Olive Ridley turtles with success. A good collection of images of stranded turtles on Cox's Bazar beach were provided by M.A. Hannan to support this research.
This project aims at developing a solution for detecting handedness of male fiddler crabs from photographs and thereby reducing human interference. At the time of writing this, the proposed solution was found working well with photos of fiddler crabs from Rezu Khal area of Cox's Bazar, Bangladesh. Next step would be testing and adapting the procedure to work with other species of fiddler crabs photographed in diverse conditions.
Procedural Content Generation
Procedural Content Generation (PCG) refers to programmatic generation of content in a virtual environment. Meaning of the terms world and content differ from one application of PCG to another, giving rise to many different possibilities. When utilized properly, I believe it can be a very good tool for studying natural phenomena.
An special class of procedural content generation and well known problem in computer graphics, simulation and modeling. My research interest for terrain generation focuses on generation of terrain with mixed geographic features and generation of parametrically controlled terrain models. Some of my published work on this field are :
- Generation of Mountain Ranges by Modifying a Controlled Terrain Generation Approach ICCIT 2008, Khulna, Bangladesh
- Parametrically Controlled Terrain Generation GRAPHITE 2007, Perth, Western Australia
- Repeated Magnification : a New Approach to Generating Artificial Mountain like Terrain GRAPHITE 2006, Kuala Lumpur, Malaysia
Part of my undergrad research works with GRAPHITE. Modeling and rendering of artificial plant is one of the most challenging fields of computer graphics with still some areas left unexplored. In our Plant Modeling project, we experiment with different plant modeling techniques as well some of our own techniques. The fractal models of plants draw most of our interest.