Project2 / GroupYohBrandonFinnBlogApril 29,2009 by bmsmith (link) I fixed our parser so it no longer hangs on some BVH files. I also now write out a slightly modified version of a BVH file that handles quaternions, and I can read those same files of course. Yoh set up a Google spreadsheet, and we've been adding/updating it every couple of days with tasks, their priorities, and who's working on each, which is perhaps part of the reason we haven't written to this blog much yet for Project 3. April 24,2009 by bmsmith (link) Are we using this space for Project 3 also? I'll post here until I hear otherwise. Our concatenate/blending code from Project 2 was not ideal, and the rotation update that would make the first frame root joint rotation of the second motion consistent with the last frame root joint rotation of the first motion didn't really work for significantly rotated motions. Therefore, I took some time and wrote a better implementation that correctly translates and rotates the second motion sequence of a two-sequence concatenate/blend. For the rotation part, the updated code takes quaternions as input. I also tried to write a function that would convert quaternions back to Euler rotations, but I couldn't get it to work. It did, however, make for an hilarious sequence in which the character appeared to be picked up by his feet and slammed against the ground a la the World Wrestling Federation. Blayne just about bust a gut laughing. April 15,2009 by bmsmith (link) Finn and I were tinkering with the project last night; we couldn't quite get tracing end effectors over time to work, but with some additional attention today it works. The traces fade as they extend away from the current position.  April 10,2009 by bmsmith (link) So I'm looking at the screen capture below... On my office computer I can barely see the checkerboard. This is different than the Storm lab computer we were using, which displayed it more clearly somehow. I think we had the right idea to make a more subtle, less obnoxious checkerboard, but we probably went a little bit too far. We should probably make the darker gray just a little darker. My fault! :) April 9,2009 by bmsmith (link) Finn, Yoh, and I worked on the project today. We got the motion working, we added functionality to a fancy-pants text box at the bottom of our window (to provide information to the user), we added a play/pause button, drop shadows, and a slider that allows the user to scrub through frames at arbitrary speed. We also improved the visual appeal of our graphics world. April 9,2009 by finn (link) We've tried roughly 12 bvh files and we have yet to find one that won't load and play. For bonus features we are thinking along the lines of fancier blending, tools to better visualize the motion and a tracking camera.  April 3,2009 by finn (link) At the moment we are just drawing cubes at each of the joints. The dark gray box on the bottom is a console to display feedback to the user.  April 3,2009 by bmsmith (link) I have the parsing done. Well, I probably need to make it a little more robust and debug it more completely, but it works on at least four random .bvh files I tried it on. I also created the structures that store the data, and those are working now. I created a class called Motion that stores a vector of pointers to the root joint for each frame. The joint objects are defined by a class called Joint. Each joint object has a vector of pointers to child joints, along with fields for offset, rotation, name, etc. Yoh has been working to generate a simple figure, and Finn has been working on the 3D world and the GUI. We'll have a screen capture of all of this soon, today. March 30,2009 by bmsmith (link) We have decided to use the following for our project: March 26,2009 by gleicher (link) Nothing Meaningful Yet |