WEBVTT 0 00:01.280 --> 00:08.810 Let's dive deeper into steganography and to see how an image can embed another file, without destroying 1 00:08.840 --> 00:13.730 the original one or without anything being noticed. 2 00:13.730 --> 00:17.300 An image is composed of pixels. 3 00:17.630 --> 00:24.710 Let's take for example an RGB image, sometimes referred to as a true color image. 4 00:24.710 --> 00:33.650 The color of each pixel is determined by the combination of red, green and the blue intensities stored 5 00:33.770 --> 00:35.480 in each pixel. 6 00:35.480 --> 00:40.340 For example this is the RGB code of this color 7 00:43.040 --> 00:47.650 and the RGB code of this one. RGB 8 00:47.660 --> 00:57.850 images are stored as 24-bit images where the red, green and blue components are 8-bit each. 9 00:57.900 --> 01:06.590 This yields a potential of 16 million colors. Usually using an image processing application 10 01:06.590 --> 01:14.600 you can see each pixel’s color as 3 numbers in base 10, between 0 and 255. 11 01:15.230 --> 01:18.560 These are the numbers for this color 12 01:18.650 --> 01:26.330 they are 255, 0 and 102. For this one 13 01:26.370 --> 01:29.200 the red intensity is 51, 14 01:29.280 --> 01:40.670 green is 204 and the blue 255. These are the red, green and the blue components. 15 01:40.720 --> 01:50.800 Note that an 8 bit number is between decimal 0 when all beats are zero and the decimal 255 when all 16 01:50.800 --> 01:52.650 beats are 1. 17 01:52.750 --> 01:57.790 These are some examples of colors and they are RGB code. 18 01:57.790 --> 02:07.030 Note that you can also represent them in hexadecimal if you simply convert these numbers to base 16. 19 02:07.050 --> 02:10.390 This is the hexadecimal code of this color. 20 02:12.310 --> 02:22.070 Let's write to the RGB code of color orange! This code! So color orange 21 02:22.420 --> 02:29.870 RGB 255 , 153 and a 0. 22 02:29.870 --> 02:31.650 This is its code. 23 02:31.650 --> 02:37.370 Now let's convert these values to binary or base 2. 24 02:37.410 --> 02:50.610 This is a very simple operation but I'll use this online tool; so from decimal to binary 255 - 25 02:50.900 --> 02:54.470 -this is the binary number - and I'll copy paste it here. 26 02:56.730 --> 03:05.330 Then we have 153. 27 03:05.460 --> 03:15.600 I'll take the binary number and 0 which means 8 bits of zero 4 6 8. 28 03:16.080 --> 03:22.420 Okay, this is the color orange in binary. 29 03:22.540 --> 03:23.900 Perfect! 30 03:23.920 --> 03:32.770 Now this is a term called the "least significant beat" or LSB of any byte and is the right-most 31 03:32.770 --> 03:35.330 bit. This right 32 03:35.360 --> 03:38.910 most one is the least significant bit. 33 03:39.140 --> 03:46.880 This one! This one for the second bite and this 0 for the last byte. 34 03:47.170 --> 03:55.750 Now in the context of steganography the secret file will be hidden by manipulating and storing information 35 03:56.050 --> 04:04.910 in the least significant bits of the image or the sound file; if a user manipulates the least significant 36 04:04.910 --> 04:14.580 bits of a color in a pixel the effect will be so small on the amount of that color that the human naked 37 04:14.580 --> 04:17.570 eye will not perceive it at all. 38 04:17.570 --> 04:26.810 However the more bits a user replaces the more obvious the alteration will appear to both a statistical 39 04:26.870 --> 04:30.270 and visual inspection. Let's 40 04:30.270 --> 04:36.660 suppose we want to embed 3 bits 1, 0 and 1 in this pixel. 41 04:40.340 --> 04:51.260 1, 0 and 1 so the resulting pixel will be this 1. 42 04:51.270 --> 04:52.620 Here is 1. 43 04:52.800 --> 04:54.510 Here I'll put 0, 44 04:54.630 --> 04:57.180 the second bit and 1 45 04:59.930 --> 05:03.380 So the new RGB code of the color will be 46 05:06.320 --> 05:11.910 the first byte will remain 255, okay. 47 05:11.940 --> 05:14.340 Let's see the value of the second byte! 48 05:17.050 --> 05:20.620 Now I'm converting from binary to decimal. 49 05:20.800 --> 05:26.380 This is the binary number and it's 152 in decimal, in base 50 05:26.380 --> 05:26.620 10. 51 05:31.180 --> 05:33.190 And the last one will be 1. 52 05:39.440 --> 05:47.570 The user may later recover the secret information by extracting the least significant bits of the manipulated 53 05:47.600 --> 05:50.910 pixel and recovery the original message. 54 05:50.930 --> 05:53.660 Let's see how this new color looks like! 55 05:59.750 --> 06:01.660 This was the original color, 56 06:01.700 --> 06:06.960 this one, and the new color is this one. 57 06:06.960 --> 06:16.580 It's the same color for the human naked eye. By the way if you modify the least significant bit it means 58 06:16.640 --> 06:25.640 you modify one bit in every group of eight bits and that means 1 divided by 8 or 12, 59 06:25.700 --> 06:35.680 5% so in an image of one megabyte you could embed a file of 125 kilobytes. 60 06:36.520 --> 06:43.870 In practice steganography tools will allow you to embed the much less information in order not to be 61 06:43.870 --> 06:44.580 detected. 62 06:45.360 --> 06:49.260 Okay, that was the theoretical part of steganography. 63 06:49.480 --> 06:57.190 Of course each tool can use a different algorithm or variation of more algorithms but the fundamentals 64 06:57.280 --> 06:59.310 are the same. 65 06:59.380 --> 07:04.540 They have to modify the cover file in such a way that no one can detect it. 66 07:04.540 --> 07:09.490 I wanted you to deeply understand how it works! In the next lecture 67 07:09.760 --> 07:14.310 I'll show you a some hands on applications of steganography. 68 07:14.350 --> 07:16.000 See you in a few seconds.