5. OpenCV Tutorial - Learn Classic Computer Vision & Face Detection (OPTIONAL)/17. Bitwise Operations - How Image Masking Works.srt

1
00:00:01,500 --> 00:00:06,900
So let's start doing some bitwise operations and include in the device operations and masking because

2
00:00:06,960 --> 00:00:10,540
that's essentially what you will be using these bitwise operations for.

3
00:00:10,800 --> 00:00:15,460
They're quite handy when you have to mask images which you all seem to want to discuss but for now we'll

4
00:00:15,480 --> 00:00:17,640
just introduce the topic selflessly.

5
00:00:17,640 --> 00:00:20,200
Let's create some chips here.

6
00:00:20,400 --> 00:00:24,160
So we're going to create a square and an ellipse here.

7
00:00:24,160 --> 00:00:29,140
Now you may be familiar with creating a rectangle or square I call it because it's same dimensions.

8
00:00:29,140 --> 00:00:33,080
I decide for this one above that an ellipse is slightly different.

9
00:00:33,090 --> 00:00:35,870
It doesn't actually follow the same standard as a circle.

10
00:00:36,270 --> 00:00:40,440
You can check the open see the documentation to get some details I won't go into it in this chapter.

11
00:00:40,440 --> 00:00:42,660
Here to take up too much time.

12
00:00:43,030 --> 00:00:46,620
This run this function and we see it's here.

13
00:00:46,980 --> 00:00:51,780
So this ellipse a single left creates and has actually not a full ellipse of the parameters to create

14
00:00:51,780 --> 00:00:55,830
sort of a semi hemisphere type image here.

15
00:00:55,830 --> 00:01:00,270
So what we're going to do now we're going to overlay these images and using some bitwise operations

16
00:01:00,270 --> 00:01:03,370
to illustrate the different type of operations that we have.

17
00:01:04,050 --> 00:01:05,530
So let's get to it.

18
00:01:06,970 --> 00:01:11,050
So that's actually run some bitwise operations on the images we just created.

19
00:01:11,440 --> 00:01:19,190
So by doing that we use these open C-v functions bitwise and bitwise OR bitwise Exel and bitwise NOT.

20
00:01:19,450 --> 00:01:24,880
If you're familiar with logic gates or just on gerneral programming you'd understand that what these

21
00:01:24,880 --> 00:01:25,430
things mean.

22
00:01:25,450 --> 00:01:28,810
However it's always good to illustrate that you know images itself.

23
00:01:28,810 --> 00:01:33,820
Keep in mind that Squire and Lipps has to be of the same dimensions here which is why we created the

24
00:01:33,820 --> 00:01:37,480
canvas initially a tree rendered between two pixels.

25
00:01:37,480 --> 00:01:38,650
So that's around us.

26
00:01:38,650 --> 00:01:43,750
So before we run this I just get a refresher of what all images look like.

27
00:01:44,140 --> 00:01:48,130
So let's run it and come on what are you doing it's going to happen.

28
00:01:48,760 --> 00:01:49,750
Exactly.

29
00:01:49,840 --> 00:01:52,190
Why would you assume this would have happened.

30
00:01:52,540 --> 00:01:57,700
This is the intersection of only those two images here and by intersection I mean is that it only white

31
00:01:57,700 --> 00:01:58,220
areas.

32
00:01:58,240 --> 00:02:02,890
And keep in mind these statements here they would when you have a binary type image either black or

33
00:02:02,890 --> 00:02:05,130
white always a grayscale image.

34
00:02:05,140 --> 00:02:07,220
It's not exactly going to look like you imagine.

35
00:02:07,240 --> 00:02:09,060
You can try it on your own.

36
00:02:09,070 --> 00:02:09,720
I love it.

37
00:02:09,850 --> 00:02:12,180
I'm eliciting the concepts of these bitwise operations.

38
00:02:12,180 --> 00:02:17,770
You see this white area here and see only parts of those two images that intersected so let's look at

39
00:02:17,770 --> 00:02:20,460
it or that wise operation.

40
00:02:21,280 --> 00:02:25,880
Exactly as we anticipated both images shown here.

41
00:02:26,200 --> 00:02:31,040
Since this sort of takes the signal and the software and we do come together here.

42
00:02:31,440 --> 00:02:33,930
So what do you think exo of to do.

43
00:02:34,550 --> 00:02:35,600
Let's see.

44
00:02:35,620 --> 00:02:37,230
EXO is sort of a weird one.

45
00:02:37,240 --> 00:02:38,800
Sort of looks like a reverse.

46
00:02:38,900 --> 00:02:40,910
So the intersection between these.

47
00:02:40,960 --> 00:02:48,640
So what this means is that anything that's an axle goes back to 0 0 0 black and red remains with the

48
00:02:49,060 --> 00:02:50,650
images exist like a wall statement.

49
00:02:50,650 --> 00:02:51,990
It shows up here.

50
00:02:52,390 --> 00:02:56,830
So it can be useful some some things you can get maybe a little and figure out what you may need it

51
00:02:56,830 --> 00:02:57,450
for.

52
00:02:57,820 --> 00:02:59,070
And others are not.

53
00:02:59,070 --> 00:03:01,270
Now keep in mind not is different.

54
00:03:01,270 --> 00:03:05,970
Not just ticks into one takes from one image into consideration.

55
00:03:06,070 --> 00:03:11,920
So it's not is actually equivalent to inverse of an image and you'll see that shortly.

56
00:03:11,920 --> 00:03:14,000
So let's look at what that is.

57
00:03:14,200 --> 00:03:15,680
So let's bring it in here.

58
00:03:16,080 --> 00:03:18,500
Memorize this one.

59
00:03:18,500 --> 00:03:25,240
So this is a square run in a bitwise NOT operation as you can see it basically includes ticklers which

60
00:03:25,240 --> 00:03:31,470
is actually quite useful in many open any functions which you will come in to come across later on.

61
00:03:31,480 --> 00:03:33,760
So that's an introduction to bitwise operations.

62
00:03:33,760 --> 00:03:35,820
Hope you found it useful.