WEBVTT

00:00.170 --> 00:00.830
In Russia.

00:01.100 --> 00:07.400
As we have seen using the TF2 library, there are two different ways to express the orientation of our

00:07.400 --> 00:09.290
reference frames in space.

00:09.970 --> 00:17.650
Initially when we approached the study of the kinematics, we used rotation matrices and so it was possible

00:17.650 --> 00:25.090
to decompose the orientation of a reference frame with respect to another into three fundamental components,

00:25.120 --> 00:30.100
three elemental rotations around the Z, Y and X axis.

00:30.830 --> 00:32.050
These three angles.

00:32.060 --> 00:39.710
These three elemental rotations are commonly referred to as Euler angles, and they allow us to represent

00:39.710 --> 00:45.530
the orientation of a body in the three dimensional space in a simple and intuitive way.

00:46.230 --> 00:53.510
However, in the laboratory lessons we also saw that rose to use another convention for the orientation

00:53.520 --> 00:56.790
representations, namely the Quaternions.

00:57.360 --> 01:03.900
We also saw that this tool is composed of four variables, which are the four variables that we had

01:03.900 --> 01:07.170
to assign to the orientation of the reference frames.

01:07.890 --> 01:14.970
These two tools serve the same purpose, representing the orientation of our reference frame with respect

01:14.970 --> 01:15.840
to another.

01:16.630 --> 01:23.530
However, in computer science, whenever it is needed to express the orientation of a body, quaternions

01:23.530 --> 01:24.610
are preferred.

01:25.060 --> 01:31.750
This is not only the case in robotics, but also in video games development, for example, in computer

01:31.750 --> 01:35.050
vision, animation and also special effects.

01:35.680 --> 01:42.550
In the upcoming lessons, we will analyze these two equivalent tools and see the reason why Quaternions

01:42.550 --> 01:46.090
prevail over rotation matrices in computer science.