using GDI, how would i detect the color of a pixel ( for colision detect)

Devide the space of the map into a grid. and check each grid position surrounding the bot for altering pixelcolours every loop. The resolution of the grid depends on the speed you want out of it. A denser grid requires a lot more checks to be done every loop and consequently slows down the program. A very course grid will give inaccurate results cos the changing colours might actually overlap a single grid lot. Matter of trial and error really. fidde with the grid resolution until your happy with the results. You can get it fairly high I think, since you will only need to check the grid spots directly surrounding the bot and not the entire grid. Note that this method will not make the bot behave very intelligently. It cannot anticipate coming obstacles until it actually runs into it. You can change this a few ways. 1. Expand the Checking radius around the bot. This way it can tell if there is a wall before it reaches it. This will still make it look like its blind though. 2. Use a pathfinding alghoritm like A* (A-Star). This is a bit more CPU intensive, but can be _very_ effective in plotting the shortest route from point A to B and avoiding any obstacle along the way intelligently. This method is often used in games for Movement AI. This method requires a bit of study and isnt always easy to implement, but its very flexible as you can expand its results to make the bot not only choose the shortest route, but for instance the safest (you might have dangerous pits and enemies on your map you wish to avoid). Hope the above helps a bit in getting this done. The method you choose will depend largely on the level of complexity you wish to achieve so what its gonna be is really up to you.

1 Reply

Know the answer? Post it — somebody with the same question will find it here.

jimteeuwen22y ago
Devide the space of the map into a grid. and check each grid position surrounding the bot for altering pixelcolours every loop. The resolution of the grid depends on the speed you want out of it. A denser grid requires a lot more checks to be done every loop and consequently slows down the program. A very course grid will give inaccurate results cos the changing colours might actually overlap a single grid lot. Matter of trial and error really. fidde with the grid resolution until your happy with the results. You can get it fairly high I think, since you will only need to check the grid spots directly surrounding the bot and not the entire grid. Note that this method will not make the bot behave very intelligently. It cannot anticipate coming obstacles until it actually runs into it. You can change this a few ways. 1. Expand the Checking radius around the bot. This way it can tell if there is a wall before it reaches it. This will still make it look like its blind though. 2. Use a pathfinding alghoritm like A* (A-Star). This is a bit more CPU intensive, but can be _very_ effective in plotting the shortest route from point A to B and avoiding any obstacle along the way intelligently. This method is often used in games for Movement AI. This method requires a bit of study and isnt always easy to implement, but its very flexible as you can expand its results to make the bot not only choose the shortest route, but for instance the safest (you might have dangerous pits and enemies on your map you wish to avoid). Hope the above helps a bit in getting this done. The method you choose will depend largely on the level of complexity you wish to achieve so what its gonna be is really up to you.
0