• Glowing Metaballs

    AS3 Metaballs

    Metaballs came around in the 90s with the Demo Effect. What made metaballs so cool was their squishy look and melding into other nearby metaballs. First I found a nice article and stuck it into as3. It turned out really slow (even after the suggested optimizations), but it worked and made some cool pictures. A metaball, according to that article, is defined by the function: F(x,y) = R / sqrt( (x-x0)^2 +...
  • Vehicle Following the Path

    A* and Pathfollowing Steering Behavior

    In the last three posts, we explored the how A* works, then we put A-star into code, then we looked at different heuristics for A*. Now we will combine A* pathfinding with the path following steering behavior. We have already done most of the work to do this. But we'll do a quick review in case you missed those posts. A* Pathfinding In A* we take a grid of cells and find a path from start to finish. Each...
  • Sierpinski triangle

    Sierpinski Triangle

    The Sierpinski Triangle is a famous fractal named after the mathematician who discovered it. It is made up of nested triangles that get smaller and smaller. If you were to zoom in, you would see the same image (though it may be translated) and you could zoom forever and never find the end. So, how do you create the Sierpinski Triangle in AS3? First you have to know how to do recursion. Recursion is...
  • AStar

    A* Pathfinding Basics

    A* (A-star) is an algorithm used for pathfinding. Pathfinding is where a computer computes the shortest (or best) path from a start point to an end point through a grid or nodes. We will be using a grid to find a path. The green cell is the start point and the red cell is the end point. The computer must find the shortest route from start to finish without going through the walls. We are using a grid because...
  • Vehicle Following the Path

    A* and Pathfollowing Steering Behavior

    In the last three posts, we explored the how A* works, then we put A-star into code, then we looked at different heuristics for A*. Now we will combine A* pathfinding with the path following steering behavior. We have already done most of the work to do this. But we'll do a quick review in case you missed those posts. A* Pathfinding In A* we take a grid of cells and find a path from start to finish. Each...
  • Concave Polygon Working

    Separation of Axis Theorem (SAT) for Collision Detection

    The Separation of Axis Theorem(SAT) is a technique to test whether two convex polygons are colliding. The SAT theorem: "given two convex shapes, there exists a line onto which their projections will be separate if and only if they are not intersecting." The line where the shapes have disjoint projections is called the separating axis. A separating line is the line that can be drawn between the two shapes,...

  • Using the Separation of Axis Theorem

    A few months ago, I posted on the separation of axis theorem. You can learn all about SAT and how it works here. What that post failed to do was use the SAT. We will explore using SAT for collision detection in this post. In the post on SAT, we created the code to detect collisions, however use of the code was not covered. (A download of all the classes used in the post is available at the bottom). The Shape...

Algorithms

Vehicle Following the Path

A* and Pathfollowing Steering Behavior

Nov 10th

Posted by rocketman in AI

5 comments

Vehicle Following the Path

In the last three posts, we explored the how A* works, then we put A-star into code, then we looked at different heuristics for A*. Now we will combine A* pathfinding with the path following steering behavior.

We have already done most of the work to do this. But we’ll do a quick review in case you missed those posts.

A* Pathfinding

In A* we take a grid of cells and find a path from start to finish. Each cell has an x,y for location, a parent for creating the path, and F, G, and H values for calculating the More >

, , , , , ,
The Diagonal Shortcut produces this path.

A* Heuristics

Nov 9th

Posted by rocketman in AI

No comments

In the first post on A*, we explored the theory behind A*. Then we took A-star and put it into code. Now we will look at different heuristics for A* and how they affect the path.

The Manhattan Method

The Manhattan Heuristic produces this path.

This is the heuristic function we used in our implementation of A*. The Manhattan method will get you to the end cell in the least steps, but it doesn’t always guarantee the shortest path. It will usually give the shortest path, but it if doesn’t, the chosen path won’t be too much longer. The Manhattan heuristic overvalues the cost More >

, ,
AStarFinal

Putting A* into Code

Nov 8th

Posted by rocketman in AI

5 comments

In yesterday’s post on the basics of A star, we explored the theory behind A*. Today we will take that theory and put it into code.

Because A-star uses a grid made up of cells, we need some classes to store that data. First the cell class:


package com.rocketmandevelopment.grid.cells {
import flash.display.Graphics;

public class Cell {
public var f:Number = 0;
public var g:Number = 0;
public var h:Number = 0;
public var isClosed:Boolean = false;
public var isOpen:Boolean = false;
public var isWalkable:Boolean = true;

These public variables are the key variables that a cell needs to have. The F, G, More >

,
AStar

A* Pathfinding Basics

Nov 7th

Posted by rocketman in AI

No comments

A* (A-star) is an algorithm used for pathfinding. Pathfinding is where a computer computes the shortest (or best) path from a start point to an end point through a grid or nodes.

We will be using a grid to find a path. The green cell is the start point and the red cell is the end point. The computer must find the shortest route from start to finish without going through the walls.

We are using a grid because it is a simple way to represent the concept behind A-Star. You can use this same algorithm to find a path through connected More >

, ,
PointInPoly

AS3 2D Ray Casting For Collision Detection

Jul 8th

Posted by rocketman in AS3

4 comments

Ray casting is a technique usually used in 3D worlds (often for rendering). Ray casting is exactly what it implies a ray is cast from a start point until it hits something (or a max length, so the computer doesn’t die). Ray casting in 2D is great for collision detections, especially fast-moving objects (bullets).

In this post, we will learn the basics of ray casting in 2D, then we will apply it to bullets in a simple environment.

A ray is just a line. It has a start point and an end point. In order to use this method effectively, we need More >

, , , , , , , ,
MetaDiamonds

AS3 MetaDiamonds

Jun 25th

Posted by rocketman in AS3

2 comments

MetaDiamonds, Click to watch!

The next MetaShape is the MetaDiamond. Its like the metaballs and the MetaEllipses but with diamonds instead.

The MetaDiamonds will be based off the filter used in the previous MetaShape posts. It will use pixelbender.

MetaDiamonds are defined by this functions:

M(x,y) = R / ( |x0-x| + |y0-y| )

X and Y are the current pixel location. x0 and y0 are the diamonds coordinates.

Now using the filter from the previous posts, it is pretty easy to put this into pixel bender

<languageVersion : 1.0;>

kernel MetaDiamonds
<   namespace : "com.rocketmandevelopment";
 vendor : "Rocketman Development";
 version : 1;
 description : "MetaDiamonds";
>
{
parameter float More >
, , , , ,
Glowing Metaballs

AS3 Metaballs

Jun 15th

Posted by rocketman in AS3

2 comments

Metaballs

Metaballs came around in the 90s with the Demo Effect. What made metaballs so cool was their squishy look and melding into other nearby metaballs.

First I found a nice article and stuck it into as3. It turned out really slow (even after the suggested optimizations), but it worked and made some cool pictures.

A metaball, according to that article, is defined by the function:

F(x,y) = R / sqrt( (x-x0)^2 + (y-y0)^2)

Where x and y is any point on the stage, x0 and y0 is the location of the metaball, and R is the metaball’s radius. In order to make metaballs, I created More >

, , , , , ,