Chris
| CS 4324 Chris |
|
Project 2 Tacker |
For this project, our team decided to use the videos of class that were provided with the project as opposed to creating our own video. The team also decided to use as much of the provided code as possible. The hardest part of the project was working with the different files of code and to figure out what each of them does. We tried to combine the ideas from the sample files and to reuse them in our code. It was also a challenge to pick the direction we wanted to pursue given that there were multiple options to try.
After exploring other options, including color recognition, the team decided to use image differencing to find and illuminate our professor. We got the idea of image differencing from /net/hb3/cga/sgi-camera/cpl which was a directory provided in the project specs, but we used code from /net/hb3/cga/sgi-movies with additions and modifications of our own. The latter code was used, since it already opened and processed movie files correctly whereas the former didn't. We chose to have the image differencing run and then to track the professor we filled in his image with white. Our solution worked well, although there are some extra white lines where the projection screens in the background are moving (in between the screens). We were able to trim those lines so they are not as intrusive now. Our code works with movies 1 and 2, but it is difficult to understand what is happening with the head cam movies since everything is moving.
One of the main reasons for choosing image differencing was that the colors on Chris were not always constant tints so color recognition was more difficult. Also, our professor tends to move around a lot during class which lends to the effectiveness of image differencing. And finally, it was just plain easier to get the image differencing to work.
There are no bugs in the code to our knowledge.
Our source code is located in /net/hc26/shleyf/prj2
When the program runs, you will see Chris' form in white moving around on the screen.