For a while I have been trying to think of ways to incorporate Rovio’s fantastic game Angry Birds into my physics class. On twitter today a tweet sent me to the Sweeny Math blog to a post about using Geogebra to analyze Angry Birds. In short Mr. Sweeny had his students analyze the trajectory of an Angry Birds launch (I think he must have done a screen shot post launch) and then let Geogebra find the curve fit of the line. Now, I have been working Geogebra this year and really find that it is a powerful program for analyzing mathematical systems, but I don’t know that it is best tool for analyzing an Angry Birds video.
For analyzing video in class (and for fun) we use a freeware program called Tracker. Tracker is a pretty amazing program especially considering software of this quality can cost a lot of money and still not be as good as Tracker is. I didn’t really have a project in mind, so I decided to jump into this analysis as an open-ended question which are becoming all the rage in science classes. So, I went to YouTube to download an Angry Birds video. (I downloaded the video using Easy YouTube Video Downloader which is a FireFox Add-On.)
The basic premise of Tracker is that you click on a spot of an object for every frame of a video and Tracker than turns the grid of the video into an x-y coordinate plane and can compute the displacement and velocity of an object. Here is how the Tracker screen looked after I finished clicking:
So, now I wanted to get some data. This is where Tracker is awesome.
So, these graphs show the shapes and relationships that we would expect from a projectile launch: Constant velocity in the x-direction and acceleration in the y-direction. However, now, I wanted to do some analysis of the accuracy of the program. So, from here I was going to treat this first video as a control and assume that it was correct. In order for Tracker to return numerical data the scale of the video has to be set so that the program knows how “large” each pixel should be.
I tried to find an object in the video that would always be the same and so I settled on one of the yellow birds. I adjusted the size of the bird until the slope of the Vy v. t graph was -9.81 which should the acceleration of gravity.
In order to test if other objects on the screen would follow the same behavior, I tracked the motion of one of the falling pigs.
So, the acceleration of the pig was the same as the yellow bird, so it seems to work out.
From here I’m not sure where I would go next, one problem I can see is that depending on the zoom of the level I don’t know if I could use the same size of the yellow bird consistently between videos. I think if I could zoom in the closest level in each video in the game then I could use the yellow bird size across videos. After writing this post, I found that A Professor of Physics at Southeastern Louisiana University tackled this same issue and came to similar conclusions, but didn’t really provide any revelations into the issue of zooming in on the bird. He did remind me Tracker’s ability to track motion of a panning video, which may help since all of the Rovio videos on YouTube start with the default zoom. He did try to include a little more quantitative analysis in his second Angry Bird’s analysis, but made the assumptions as before. (In case you haven’t seen enough, here is a similar analysis of Super Mario Brothers.)
The next experiment I want to try is to determine the mass of the birds. This video wouldn’t work well for it, but if you could assume the density of the rock barriers and then find the volume (find the height and width using the scale and the assume the depth) then you could find the mass of the rock barriers and using the conservation of momentum you could find the mass of the bird. I am sure there are other ways to go, but that would be a fun start.
Here is the final result: