Link: https://informationalmind.appspot.com/2011/10/6/AfterWatchingComputerGraphicsCourse.html
I have watched a course video of Computer Graphics(CG) these days.
Course link first: Graphics
Why this?
Well, when I have read a book of CG(this one, translated version of this one), I wanted to find a course video to revise some basic knowledge of CG. So I searched on YouTube and found a series of videos of CG course - the course I have watched. Following the link of the uploader, the web site of this course was easily found.
While searching course videos, an Indian course is the first result, but that series of videos are hard to catch for me because of the pronunciation...
This course was given by Utrecht University, which I have not heard before. Sorry for my ignorance.
This is a very basic CG course that targets to undergraduate, and no advanced math knowledge is presumed(linear algebra, for example). It starts from a relatively low level, and covers math that related(vectors, matrices and related operations), transformation, texture, shading, operations used in graphic pipeline, and some little advanced topics like shadows, ray tracing, etc..
The explanation of each step of this course is very clear, not like some math courses I had in university – theorem, proof, theorem, proof... It goes ahead, when problems occur, stops to solve it in a really intuitive and easy to understand way, then goes on. You can know what is there, and how to solve it, and why it works. Compare to some of the math courses I have had that knowledge are only be given but never explained where it can be used and how it can be used, this may be a more effective approach to master something.
The most gains for me from this course may be the matrices used for doing transformation, especially the matrix used at perspective projection step. When I read the CG book motioned above, the perspective projection matrix is presented as a single step, a very complicated step. But in this course, it is taught in two steps and this make things pretty easier. The first one is to transform a view frustum into an orthographic view volume(the one used for parallel view), and then transform orthographic volume into a canonical view volume(a box which ranges from –1 to 1). In this way, it gets rid of the aspect ratio and view port width be considered in the camera space to orthographic volume step and these are relatively easy be dealt with in orthographic volume to canonical view step. Furthermore, I understand what is the affine transformation when perspective projection was taught.
Texture, culling, clipping, z-buffer, rasterization and shading are something I already understand. Finally, a little about shadows, radiosity and a lot of ray tracing.
After all, I have to say: I love learning by videos~ We can stop at any time and replay whatever part we want in this way.
