Compression of similar images and cloud storage.

Introduction

I was thinking about a new startup plan that will make my pockets filled. But really got disappointed of my damn small brain's cheap ideas those were already in the market. Every ideas I got, it is listed somewhere in the Google with some big firms. But finally I ended up with an idea inspired from the very famous show 'Silicon Valley' that everyone must have thought about it once at least.
But it's not like the bloody damn internet breaking algorithm where we can compress a big 12 Gigabytes file into 100 Kilobytes without any loss. Nah, we need a couple of decades to solve that.

The idea

What I thought about is, what if we have a hundred images with some of its parts looking similar each other. let's imagine the 1/4 of every picture looks the same and we can take that part as a common image. It would buy us 25% of the space for free. This is the least-cool idea came up in my mind.
When we build a cloud storage and millions of users upload their images, there would be at least a 100 K image pairs look similar each other. And we can save a good amount of space.

The relevance

I am not sure whether the popular cloud storage services like Google drive, Onedrive or DropBox are using this concept or not. But it's sure that if we can develop an algorithm, it would be a big milestone.

References

I found only a few articles about the image set compression and only one impressed me.
This is the link I found in the internet.
It says we can build an algorithm to compress a set of similar images with priorities.
These are the steps of the algorithm:

Conclusion

It's a pleasure if anyone is interested in such a project. Do comment your opinions. Let's help each other to bring more articles and build a good algorithm. Also we have to build a cloud storage. Let me know if you are interested.