Jump to content

  • Log In with Google      Sign In   
  • Create Account

#ActualKnolanCross

Posted 11 July 2013 - 01:10 PM

This link has a full explanation of how Navmeshes are created and used:

http://www.critterai.org/nmgen_study

 

 

Edit:

Read your question wrong, sorry. The link should help you.


#2KnolanCross

Posted 11 July 2013 - 01:03 PM

A* is an algorithm that will search the shortest path in a weighted graph (and that is all). When you use a navmesh you will have a graph where each of the neighbor regions of a region are the adjacentes nodes of a node in a graph (so basically, the neighbor nodes are the candidates of your open list). In other words navmesh is just a way to represent a graph that the A* will use.

 

Please notice that A* will not give you the navigation path, it will only tell you that the shortest sequence of edges to go from one node to another. AFAIK to navigate people cast a ray to the border of the region that conects your current node with the next node you have to go and follow this line.

 

This link has a full explanation of how Navmeshes are created and used:

http://www.critterai.org/nmgen_study

 

Using navmesh pathfinding is pretty advanced, maybe you may consider using a simpler representation for a start, such as grids.


#1KnolanCross

Posted 11 July 2013 - 01:00 PM

A* is an algorithm that will search the shortest path between nodes, when you use a navmesh you will have a graph where each of the neighbor regions of a region are the adjacentes nodes of it in a graph.  So basically you the neighbor nodes are the candidates of your open list.

 

Please notice that A* WILL NOT give you the navigation path, it will only tell you that the shortest paths is given nodes and weights.

 

AFAIK to navigate people cast a ray to the border of the region that conectes your current node with the next node you have to go and follow this line. This link has a full explanation of how Navmeshes are created and used:

 

http://www.critterai.org/nmgen_study

 

Using navmesh pathfinding is pretty advanced, maybe you may consider using a simpler representation for a start, such as grids.


PARTNERS