Shortest path algorithms in a heatmap.

I have created a heatmap that shows a map or structure of my apartment. I have created it by using 24X24 grids for that i have created a dataset myself for simulations in which each grid value represents the probability of human presence in that node. For visualization consider a node represents a square of 0.5 square meters. Now i have a situation in which i will have a starting node and a destination node. I want an algorithm that can create or tell me shortest path for the 2 nodes. That algorithm has to consider the facts that some nodes will have 0 probablilty for the areas of the heatmap representing furnitures. Any suggestions regd the algorithm with some details too.

[Matt J]

How do you define the distance between two nodes? How are the probabilities related to the selection of a path?

[agrim agarwal]

So distance can be represented or measured by number of nodes we have to travel. As far as probability is concerned it is not the best word to explain the scenario. Consider value of node as more the value is more are the chances of presence of our subject on that node. So lets say a PIR sensor is tripped and it's circle covers around 4 such nodes then the one with the highest value/probability can be assumed to be the node with presence of the subject.

[Matt J]

So distance can be represented or measured by number of nodes we have to travel.

In that case, the straight line between the two nodes is the shortest path. If the problem is to be non-trivial, there need to be some constraints on the selection of the path. You haven't mentioned wthat those constraints are.

[agrim agarwal]

those constrains are the nodes representing furnitures and other inaacessible parts of the taraget apartment.

Answer 1

[Image Analyst]

See Steve's blog: http://blogs.mathworks.com/steve/2011/11/01/exploring-shortest-paths-part-1/