For my undergraduate thesis, I worked on weight estimation problem in poultry houses, with Colin Usher in FPTD at Georgia Tech Research Institute.

The initial stage involved collecting 3D scans of various birds to find a correlation between a bird’s weight and volume. Methods such as ICP (Iterative Closest Point) from PCL (Point Cloud Library) and sphere fitting in MATLAB were employed to prove this theory.

ICP and maximum size sphere fitting to correlate volume and weight.

The problem with using 3D reconstruction softwares like KinectFusion, is that they are designed for modeling static scenes. This makes it harder to model chickens. I am working on an algorithm that would undo this motion to robustly generate 360 degree view models.

I have been awarded the President’s Undergraduate Research Award for Spring 2018 to continue this work. Moreover, I presented my work at the 13th Annual Undergraduate Research Symposium at Georgia Tech and was awarded the 3rd Best Oral Presentation under the College of Engineering.