Doing it MY way!
When you build a mosaic from tons of pictures, you have to downscale the images before you can use them. This naturally means loss of information in the picture, when going from say 3200×2600 to 80×60 pixels. If you have a family photo, you wouldn’t be able to recognize the people in the photo any longer.
So what do you do then?
Just passed my MCTS exam in SharePoint 2010…. Nice 🙂 It was actually quite hard
The new HTML specification, version 5, is great! There is a host of new things to explore including embedded video, SVG, Canvas element (more on that later!) and… Geolocation services! I would like to give you a small demonstration of the API.
Just spend a few hours getting Windows 7, Apache, PHP and MySQL up and running.
It should have been fairly easy, but there was a small problem.
Yes you would, and you should!
Explanation follows shortly. Over the last few weeks, I’ve been helping out a friend with his new site. He payed some programmers from India a few hundred euros to complete the site, and they did.
However, they made more than a few errors!
Going through my “huge” archive of digital camera images, I noticed that I had some duplicate images. The filename differs but the content is the same. Now I could just download a tool from the almighty internet, that could compare all my images, but that would be going to far! Why not give it a go myself?
Just about one week ago, my baby son was born so blogging was moved down the priority ladder. I honestly thought that I would have just a tiny bit more time on my hands, since he is supposed to sleep most of the time. Boy was I wrong 🙂 Anyways, the misses and I are … Read More “Family increase” »
I just read this very interresting article on Sgt. Conker on HLSL (High Level Shader Language). It’s a beginner article, but it explains very well how shaders work, and how they came to be! I can strongly recommend this article to people new to XNA and/or shaders!
Sorry… I can’t show you the mosaic since it contains 100% private pictures. Maybe I’ll go download a couple of hundred pictures from Flickr and show you! I just don’t have the time right now.
Anyways, I quickly spottet a small “error” with my algorithm. Lets say you have a source image of a clear sky. After you divide the image into a grid, and scan the cells, it’s more than likely that the average color of all the cells will be the same.
First approach is to index all my photos (20.000+) to determine image resolution and average color value.
Before the average color value is calculated, the image is re sized to a width of about 50 pixels. Aspect ratio is maintained. This will improve scanning speed, as fewer pixels will be scanned, and also improve color correctness as only the visible pixels will be calculated. Once this process is done, the image is stored in a SQL Compact server database (filename, width, height, colorvalue).