0:03

Okay, now you know the big picture of steganography as well as the basics

of binary math.

Now it is time to look into putting all of this together.

Let's suppose you wanted to hide this top image, which has a secret message

inside of this bottom image, which is a picture of a galaxy.

You have already seen how to take the eight bit numbers which represent

one of the color components, red, green, or blue, and

hide the most significant bits of one in the other.

You take the most significant bits from the image you want to show up and

combine them with the most significant bits of the image you want to hide.

You've already seen this math to make this work, but

there's a bit more to this than just doing math on one pair of numbers.

You need to do this math for each pixel.

Iterating through each pixel in an image should be quite familiar by now.

And you need to do that math for the red, green, and blue components.

Since you need to do the same thing three times,

you might think about pulling that computation out into a function,

allowing you to call the function, rather than repeating the code for that task.

1:10

Speaking of functions, since this task is a bit more complex,

you might find it useful to break it down into a couple functions.

We're going to walk you through a couple functions that we broke

our implementation down into for hiding.

First, we made a function called chop2hide which is going to take the image

we want to show up in the end and produce one that looks quite similar.

It is going to do the math on each color component of each pixel

that keeps the most significant bits and zeroes out the least significant bits.

The algorithm for this function would be to iterate over each pixel in

the image and do this math to the red, green, and blue components.

You want to divide by 16, take the Math.floor of that result,

then multiply by 16.

You might want to extract this repetitive math out into its own separate function

then just call it in each place.

The resulting image should look pretty much the same as the original

as we show here.

If you were implementing, testing, and

debugging this code, breaking it into small pieces like this would be helpful.

You can check the result of this one piece and make sure it works before moving on.

2:22

The second function you might write would operate on the image you want to hide.

We call this shift because it will shift the most significant bits over

into the least significant bit positions.

That is, you would take a color value like 10110010,

and convert it into 0001011.

The 1011 has been shifted from the most significant bits to the least

significant bits.

2:54

The algorithm for this would be to iterate over each pixel and set the red, green,

and blue to the Math.floor of the original value divided by 16.

If you look at the resulting image here, it would look plain black.

That is actually a great thing since all the information is in

the least significant bits.

And the most significant bits are supposed to be all zero.

Again, you could test this separately before you move on.

3:21

The final function, called combine, takes two images and adds red to red, green

to green, and blue to blue to produce the color values of the resulting image.

It should look basically the same as the original image.

Once you have the ideas for these three functions,

your high level algorithm to hide one image in another, looks like this.

You just call the functions that we discussed and they do all the work.

Having seen the algorithms for

each of these functions, you could translate them into code.

For example, here is the chop2hide function.

You can see here, where we decided to pull out the actual math into it's own

helper function and call it for each of the red, green, and blue components.

The other functions to hide would be similar.

As we discussed, you iterate over the pixels of the image and

do the math we previously described.

What about extraction?

How do you get the hidden image back out?

We're going to leave that to you.

You've already seen the math required.

So now, you can use the seven steps to develop an algorithm and

translate it into code.

Start out with step one and work on an image with two pixels.

Write down numeric values for the red, green, and

blue, then extract the hidden two pixel image.

Once you have done that, work through steps two,

three, and four to develop your algorithm.

Then translate it into code, test it, and debug it.

Happy coding!