1
00:00:04,790 --> 00:00:05,090
Hello.

2
00:00:05,780 --> 00:00:07,670
So we are finally ready for the good stuff.

3
00:00:08,270 --> 00:00:11,990
That means that we are ready to implement the steps necessary to create the mess.

4
00:00:12,000 --> 00:00:12,710
Solving what?

5
00:00:14,100 --> 00:00:17,850
So if you remember, the very first step was localization.

6
00:00:18,360 --> 00:00:23,610
Now, localization in robotics is essentially extracting the robot boss in the given environment.

7
00:00:24,390 --> 00:00:30,450
For example, if you were in a maze, that would mean extracting the robot position and the robot orientation

8
00:00:30,570 --> 00:00:31,620
at any given moment.

9
00:00:32,750 --> 00:00:34,040
For a robot navigation.

10
00:00:34,280 --> 00:00:36,170
This becomes an important precursor.

11
00:00:36,350 --> 00:00:41,690
As without the robot tools, we would be unable to make decisions about further action.

12
00:00:43,360 --> 00:00:49,330
This also appeals to our second point that we not only to localize the robot, but we also need to do

13
00:00:49,330 --> 00:00:53,830
it at each frame to get the most updated information at all times.

14
00:00:54,980 --> 00:01:01,100
So one question that might come into your mind is, isn't this what I'm supposed to do and why aren't

15
00:01:01,100 --> 00:01:02,000
we talking about it?

16
00:01:02,810 --> 00:01:04,370
Well, yes and no.

17
00:01:05,350 --> 00:01:05,740
Yes.

18
00:01:05,920 --> 00:01:12,190
SLAM Which is simultaneous localization and mapping does in fact localize the robot at each iteration,

19
00:01:12,730 --> 00:01:19,000
but it also creates the map of the environment in the event where there is no global map available.

20
00:01:20,540 --> 00:01:27,680
The reason we have not opted for Islam and instead divided the localization and mapping into two separate

21
00:01:27,680 --> 00:01:33,830
category is because we have the luxury of observing the myth or the complete myth at all times.

22
00:01:35,050 --> 00:01:43,600
So it's like even though being very powerful is right now an overkill and has also been skipped intentionally

23
00:01:43,600 --> 00:01:44,470
for simplicity.

24
00:01:45,740 --> 00:01:51,350
So this begs the question, how do we propose to perform localization for our kids?

25
00:01:53,360 --> 00:01:56,240
Now for that, we need to look at what our case is.

26
00:01:57,450 --> 00:02:04,530
We have chosen the simplest case of an overhead vision system, be it a drone or satellite, observing

27
00:02:04,530 --> 00:02:05,880
the myth at all times.

28
00:02:06,920 --> 00:02:11,210
This means the only moving object will be our mere solving bot.

29
00:02:12,110 --> 00:02:13,310
So this is our assumption.

30
00:02:13,610 --> 00:02:17,870
And of course, a dynamic environment would have been a lot harder to solve.

31
00:02:19,100 --> 00:02:24,980
So the goal for this localization module is to localize the robot at each iteration.

32
00:02:25,160 --> 00:02:30,320
And by localizing, we mean extracting the robot position in the duty man's image.

33
00:02:31,160 --> 00:02:32,030
So let's do that.