The Manhattan distance is the distance between 2 points traveled by a taxi when he is in a city arranged in grid as:
d(A,B) = |XA - XB| + |YA - YB|
Same as Manhattan but the distance is weighted by 3. This technique is faster but gives a less precise result in number of strokes as:
d(A,B) = 3 * (|XA - XB| + |YA - YB|)
Same as Manhattan3 but weighted by 8. This heuristic is better with large maps but can miss the solution on small maps as:
d(A,B) = 8 * (|XA - XB| + |YA - YB|)
The Euclidean distance is the "ordinary" straight-line distance between two points as:
d(A,B) = √((XA - XB)^2 + (YA - YB)^2)
Dijkstra is an heuristic that check all the possible strokes, depth by depth. It will be the slower heuristic but it will always give the smaller strokes result