1 00:00:01,350 --> 00:00:08,730 And yeah, CV2 installed and I what I can with CV2 among many other things what I can do the 2 00:00:08,850 --> 00:00:14,620 very first thing is to load an image in Python from using CV2 methods. 3 00:00:14,730 --> 00:00:22,080 So I will be loading that image which is a very small image and by the way it only has 15 pixels. You know 4 00:00:22,080 --> 00:00:29,550 a normal image that you use to take pictures on your phone or whatever it has like a thousand by a thousand 5 00:00:29,580 --> 00:00:33,450 pixels which is one million pixels more or less. 6 00:00:33,840 --> 00:00:38,290 But we should keep things simple so you understand what is going on. 7 00:00:38,310 --> 00:00:41,270 CV to import that. 8 00:00:42,170 --> 00:00:44,200 Yup, that was successful. 9 00:00:44,230 --> 00:00:51,150 Now I'll create a variable and call this im for image and for gray. 10 00:00:51,540 --> 00:00:56,720 You can call it whatever you want. And CV2 to get an image to load it in Python. 11 00:00:56,730 --> 00:01:04,740 You want to use image read method and than you pass the name of your image. You can find this image in the 12 00:01:04,740 --> 00:01:05,900 resource section. 13 00:01:06,270 --> 00:01:07,810 So just go ahead and download it. 14 00:01:07,900 --> 00:01:12,840 And now this is here in the same directory with my Jupiter Notebook. 15 00:01:13,050 --> 00:01:19,260 So I can go ahead and point to the name of the file including the extension smallgray.png. 16 00:01:19,650 --> 00:01:22,080 And here is image read method 17 00:01:22,080 --> 00:01:31,170 also expects one other argument from you that can be either 0 or 1 so 0 means you want to read the image 18 00:01:31,200 --> 00:01:40,080 in grey scale and if you pass the 1 you are reading the image in BGR which means blue green and red. 19 00:01:41,040 --> 00:01:43,920 And you'll understand what each of the versions mean. 20 00:01:44,070 --> 00:01:46,990 So let's do this all first. 21 00:01:47,020 --> 00:01:51,250 Yeah I'll print that. 22 00:01:51,540 --> 00:01:52,420 And yeah, that was it. 23 00:01:52,500 --> 00:01:57,000 So what we got is a Numpy array. It's a two dimensional array. 24 00:01:57,270 --> 00:02:06,420 So it has dimensions 3 by 5 which is the same as our image, three rows and five columns. 25 00:02:06,420 --> 00:02:12,510 Now the reason we pass 0 there so we are reading that as a grey scale image is because you know this 26 00:02:12,550 --> 00:02:13,380 png file 27 00:02:13,380 --> 00:02:15,770 in reality it has three bands. 28 00:02:15,780 --> 00:02:21,740 Or you can say three layers and each layer has a certain amount of color. 29 00:02:22,530 --> 00:02:26,150 So yeah, the first layer has a certain amount of blue. 30 00:02:26,190 --> 00:02:29,110 And the second layer has a certain amount of green. 31 00:02:29,130 --> 00:02:31,990 And the last layer has a certain amount of red. 32 00:02:32,130 --> 00:02:38,870 So combining each pixel is a combination of all the intensity of blue green and red. 33 00:02:38,940 --> 00:02:44,420 I know in other programs they use RGB and OpenCV uses BGR. 34 00:02:44,960 --> 00:02:50,910 Now practically this file here, this image practically is a gray scale image. 35 00:02:50,970 --> 00:02:56,660 However each of these pixeles has its own band so the combination of them always creates 36 00:02:56,660 --> 00:03:01,080 in this case lays gray or white and black colors. 37 00:03:01,530 --> 00:03:07,140 But if we had a color image this would convert it to a grayscale, this function. 38 00:03:07,410 --> 00:03:18,170 So again what we have here is you know 187 is the value of the intensity for this first pixel. 39 00:03:18,200 --> 00:03:24,760 And then you have the other number and for white you see what white here? I think we have three white pixels. 40 00:03:24,850 --> 00:03:27,960 Now those are 255. 41 00:03:28,200 --> 00:03:39,330 All of them. And yeah, colors range from zero. Zero would be black pitch black and 255 for white. And everything 42 00:03:39,330 --> 00:03:40,030 between them 43 00:03:40,050 --> 00:03:42,210 is in the gray scale. 44 00:03:42,300 --> 00:03:50,990 If I change this to 1, I'll get a three dimensional array where this represents the blue layer the 45 00:03:51,030 --> 00:03:52,780 green and the red. 46 00:03:52,820 --> 00:03:54,120 Or you can even print that out. 47 00:03:54,120 --> 00:04:00,480 48 00:04:00,480 --> 00:04:03,770 Be careful though that this hays a bit transposed. 49 00:04:04,170 --> 00:04:10,150 So the matrices are three columns in here and the rows are vertical. 50 00:04:10,440 --> 00:04:16,080 And basically what this say is this pixel has this corresponding values. 51 00:04:16,620 --> 00:04:19,160 So it has an intensity of 187 52 00:04:19,270 --> 00:04:29,400 for blue 198 for green and 209 for red and yeah that's how you create Numpy arrays out of images. 53 00:04:30,330 --> 00:04:38,850 But how about the other way around, so creating images out of an Numpy array? Well to do that you'd use 54 00:04:38,880 --> 00:04:43,930 CV2 again .CV2 and then you would point to the 55 00:04:43,970 --> 00:04:46,230 imwrite method so image write. 56 00:04:46,590 --> 00:04:51,640 And then you want to pass a name for your image. 57 00:04:52,010 --> 00:04:56,160 Let's say newsmallgrey.png. 58 00:04:56,630 --> 00:05:07,620 And of course you need to pass your Numpy array. So let's pass image gray. Execute that, you'll get two. 59 00:05:07,710 --> 00:05:17,470 Which means that your file should have being created now. And yeah, new small gray and there's the file. 60 00:05:17,470 --> 00:05:17,950 61 00:05:18,660 --> 00:05:22,640 And I hope that strengthen your knowledge about Numpy. 62 00:05:22,650 --> 00:05:26,220 Now we can go ahead and do some more things here. 63 00:05:26,490 --> 00:05:34,970 Specifically the next lecture we will learn how to do indexing and slicing and iterating of Numpy arrays. 64 00:05:35,000 --> 00:05:35,550 65 00:05:36,090 --> 00:05:39,360 So basically you learn how to access these values. 66 00:05:39,360 --> 00:05:40,650 See you there in the next lecture.