I've been using the maze_solution function from Image Analyst http://www.mathworks.com/matlabcentral/fileexchange/27175-mazesolution for a while without problems. There are a few limitations: the maze has to be perfect and without circular paths.
That being said, it generally works very well when I test it, let me provide you with two examples of outputs:
maze-1 -> solution-1 (Clearly works well)
maze-2 -> solution-2 (Not so well)
Now let me put some rules that are not obvious about my mazes:
There are no circular paths (there are some that are trapped in the walls, but no one that the the maze solver would run into).
They begin always at the top left, and then there are four exits in the same coordinates every time.
There is always only one exit.
So, what I would like to do is, let's consider the first screenshot. It works well and 'finds' the exit, is there any way to make mat lab pop up a messagebox (using msgbox(), for instance) and say something like "Hey user, I found the solution! It's A!"? I already thought about this for a long while, but found no way of doing this. One of the solutions I thought about was, in pseudocode:
if CertainCoordinate = red pixel
return A
Whereas CertainCoordinate Could be the unchangeable coordinates (x and y) of A, B,C and D (Then I would use 4 'ifs'), but I don't really know how to implement even that. Any ideas or... something to point me in the right direction?
So, summarizing it: I have an algorithm that right now generates a red path to the exit (and I am open to better algorithms, if you have any suggestion) but my goal is to make matlab tell me what he found, instead of showing me in a image. So for instance, in the first image I would like it to open a MessageBox and say "Hey user! I found the exit, it's D!", instead of showing me the image with the red path on it. The problem is, I don't know how to teach MatLab 'where is D' and to make him recognize that he found 'D'. So, are there any suggestions on how to do that?
Thanks in advance!
Suppose you have an image called maze aset of possible x coordinates of the exit and the corresponding y coordinates of the exit. Also you can check what color the maze solving program uses, lets say this is Red.
Now the solution is quite simple. First run the maze solver, then check the following
isRed = maze(x,y) == Red;
exitxCoordinates = x(isRed)
exityCoordinates = y(isRed)
This gives you the x and y coordinates. (If no solution is found they are empty).
From here it should not be too hard to connect them to one of the letters that you specified.
Related
I'm working on a game that will always be an n x n grid. I'd like to be able to always generate a new start position from the outer edge and have a path to the center cell. My issue is that I want to be able to specify the path length instead of choosing the shortest path. Does anyone have an idea of how to do this?
I've looked at several CompSci forums, but unfortunately I'm unable to make heads or tails of what they're saying.
https://cs.stackexchange.com/questions/44401/what-algorithm-to-use-to-generate-random-path-of-given-length
In order to avoid the hardcore math, I would suggest to always use the shortest possible path and randomly generate and add some pieces to it to make it the length you want. It would not be absolute random, but for sure it should look good enough. Although it depends on your needs.
I have the following problem and I am a bit clueless how to tackle it as my programming skills are very elementary ( I am an engineer, so please dont bite my head off).
Problem
I have a point cloud, the picture above displaying one level off it. Every point is a centroid off a block (x =5, y=1, z=5) and is specefied by carteisian coordinates.
The centroids further have two values: one called "access" and one "product". If the product value is positive and pays for the access to the point I want to include it in my outcome. The red marker in the picture represents a possible starting point.
Starting Idea
As a start I am trying to set up an algorithm, that starts at the red marker, runs through the blocks left and right (along the x-axis), checks until where it would be feasible to access (so sum "product" > sum "access") and then goes to the next point (in y direction from marker) and does the same until the end of the level.
Final Goal
My final goal is that I can Flag points as accessed and the algorithm connects profitable "products" (so products that would pay for their access) on the shortest way to the access point (by setting blocks/points on the way to accessed).
I know this is a very open question and I apologize for that. I am just lacking a good starting point programming wise. I was thinking of knnsearch, but I am not sure if this is the right way to go as the blocks have different sizes and i technically want the nearest neighbour in every direction but also only one per direction.
Another idea I had was using shortestpath or creating a travel salesman problem out of it, but I am not sure how to properly implement it.
If you have any ideas or you could offer any help I would very much appreciate it. If any more information is needed I gladly provide it.
In a compute shader (with Unity) I have a raycast finding intersections with mesh triangles. At some point I would like to return how many intersections are found.
I can clearly see how many intersections there are by marking the pixels, however if I simply increment a global int for every intersection in the compute shader (and return via a buffer), the number I get back makes no sense. I assume this is because I'm creating a race condition.
I see that opengl has "atomic counters" : https://www.opengl.org/wiki/Atomic_Counter, which seem like what I need in this situation. I have had no luck finding such a feature in either the Unity nor the DirectCompute documentation. Is there a good way to do this?
I could create an appendBuffer, but it seems silly as I literally need to return just a single int.
HA! That was easy. I'll leave this here just in-case someone runs into the same problem.
HLSL has a whole set of "interlocked" functions that prevent this sort of thing from happening:
https://msdn.microsoft.com/en-us/library/windows/desktop/ff476334(v=vs.85).aspx
In my case it was:
InterlockedAdd(collisionCount, 1);
To replace
collisionCount++;
And that's it!
I have a CT scan for an heart and I am designing a device that rests on top of it. As such, getting the right lengths for certain attributes is important. The CT scan was segmented in MeshLab and my advisor gave me code that uses PLY_IO to read the ply file exported from MeshLab. From this, I have a map of the surface. surf(Map.X, Map.Y,Map.Z) outputs the 3D model. Now, what I would ideally want is to be able to select points graphically via the figure window and have Matlab either tell me what the points are or allow me to draw a geodesic line to determine its length. Question: Does anyone have any idea of how I could do this in a simple way?
Ultimately, just drawing on the figure might be ok too if I can just get it in the right orientation. Ideally, though, I would select the start and end point and then Matlab would graphically show a geodesic on the surface that I can later find the length of. I'm willing to do some programming for this, but hopefully there's something out there you guys might already know about.
One way to interactively extract points on a surface is to use datacursormode. Here's a simple example of how to get two points:
surf(peaks);
dcm_obj = datacursormode(gcf);
set(dcm_obj,'DisplayStyle','datatip',...
'SnapToDataVertex','off','Enable','on')
disp('Select first point then press any key')
pause
c_info{1} = getCursorInfo(dcm_obj);
disp('Select second point then press any key')
pause
c_info{2} = getCursorInfo(dcm_obj);
Note that if you (or the user) changes mode (e.g. by clicking the rotate button) in order to select the point, you will have to switch back to datacursor mode to move the datacursor again:
You should now have c_info{1}.position and c_info{2}.position which are two points on the surface. Calculating the geodesic is another matter - have a look on the File Exchange, see if there's anything around already that will do the job for the type of data you have already.
I'm just beginning with Image analysis in MATLAB.
My goal is to do an automated image segmention on images of plant leaves.
I have had reasonable success here thanks to multiple online resources.
The current objective, the reason why I'm placing this question here, is to able to place 25 equidistant points along each half of the margin/outline of leaf, like described in following image:
For the script to be able to recognize each half of the leaf, user can put two points within the GUI. One of these user-defined points will be on the base of leaf and the other on tip of leaf. It would be even better if a script would be able to automatically recognize these two features of the leaf.
For the output, I would like a plain text format file containing image coordinate of each point.
I'm not asking for a ready made script here, but looking for a starting point.
One way I think this can be done is by linearizing/open up the outline in such a way that it becomes a straight line. This can be done by treating any of user placed point/landmark as breakpoint. Once a linear outline is obtained it can again be broken into two halves at other user defined point and now points can be placed. One point to bear in mind here is that the placement of points for each half should start from the end that corresponds to the same breakpoint/user-defined point in each half. Now these straight lines can be superimposed on original image for reconstruction.
Thank you very much.
Parashar