If can be so kind I am looking for ideas and advice.
If you dont want to read the background please go to theFor this version
I have an idea an the approach I think will work
but don't know enough about WebGL to know if I am right.
- We have very large satellite images, 30,000 x 20,000 in size of total area.
- On top of this image we want to place and 3D objects which move over this image.
This app already exists in 2D using WPF.
I created my own version of Deep Zoom to handle this image display.
I did this by creating several pre-scaled tile folders.
Each folder contains a certain zoom level.
Each tile currently is at 512x512 but can be altered to any square sized boundary.
I have only 5 zoom levels I currently use. I use the video cards scaler to deliver all the tween sizes. Example. The user is at a zoom level of 25% I grab the tiles from the 20% folder and zoom them up 5%.
Also when the user is at zoom level "X" I only use the image tiles necessary
to cover the display at that zoom level.
In other words, I load as few tiles as possible to fill the screen with some overlap. If the user moves left or right up or down I load new tiles as necessary to display the image. This keeps ram usage low and image load times quick.
The tiles are currently 512x512 pixels in size though they can be made at any size but 512x512 seemed to perform OK.
Then application then overlays and moves several types of 2D sprites on top of the image. Thats a quick run down on what we currently have. For this version:
We want to move to the web. We want to keep the image quality if we can but want to enhance the app to have a 3D overlayed objects.
We are looking to the future so we are looking at using HTML 5
I started thinking we could use a canvas and as the user moves we reload the tiles
just as before but that does not give us 3D.
What I think is the background would be a texture but there is so much I dont know.
So I have a large question.
For a background texture can I continually swap it out and zoom and offset it based on zoom level and or offset position?
I see several examples where you load a static texture and the camera view does the reset but hey this texture is 30,000 x 20,000 pixels at 100% so that standard thing is out.
Idea's help suggestions.