I competed on a team of 4 in MIT's Mobile Autonomous Systems Lab (MASLAB) competition. In one month, we designed, built, and programmed an autonomous robot that could navigate, avoid obstacles, collect balls, and sort them by color. Code for the robot (both the firmware and the higher-level processing) is hosted on my Github: https://github.com/araju/maslab-2014. For more information about the competition that year, visit http://maslab.mit.edu/2014/wiki/Maslab_2014. To see more information about our robot, check out our team's summative report.
Chase Lambert and I created a virtual ruler that could measure objects in images taken with your camera phone if you place a credit card next to it for scale. The ruler automatically finds the credit card and allows the user to draw out the dimension he or she wants to measure. The project is still in a pre-alpha stage, but initial results are promising. Code to be hosted on Github soon.
GIFs are incredibly popular on the Internet, but they are actually an archaic and inefficient image file type. For my final project in 6.344: Image Processing, I designed a new image coding standard called JIF (just to be confusing). JIF achieves better compression than standard GIF without needing a limited color palette. Below, the first Joker is a GIF, and the second is a JIF. Notice how the JIF is smoother and has better color clarity. Check out my report for an analysis of the techniques I used.
As part of the MIT class 6.141, taught by Prof. Seth Teller, four teammates and I built an autonomous robot that could find and stack colored blocks while avoiding obstacles. Throughout the semester we also implemented different canonical robotics algorithms for areas like localization, mapping, and path-planning. More information on what my team did can be found at the team's wiki page. Unfortunately, MIT certficates may be required to view the page.