I have a problem that reminds me of Voronoi, but I'm hoping that my variation will allow me to avoid using the Voronoi algorithm, and write something quicker.
Here's a horrible image I made in Paint to illustrate my problem:
Say I have an area of a map. Each dot represents a shop. Each square represents a neighbourhood. The voronoi diagram shows the areas closest to each shop.
If one of those areas dominates a square, then that whole square belongs to that shop.
Is it possible to determine which squares belong to which shop, without having to calculate an intermediate voronoi diagram? It seems as though, since this is like a very rough approximation of a voronoi diagram, there should be a super fast shortcut to generating it.
Perhaps I'm misunderstanding, but can't you just find the vertex which is closest to the centroid of each square?
#user2615897 points out that this isn't generally correct (see comment). None-the-less, I think it would be a good approximation for a grid which looks like your example (specifically: roughly equal-area cells, with spacings comparable to the square-sizes).
My intuition is that without explicitly constructing the diagram, any approach will only be an approximation... but I'm not sure.
This (segment) of a configuration illustrates the point:
The red vertex is nearest the center of the central square, while the green-vertex owns the most area.
Related
I'm neither a geometry student or a native speaker, so apologies if my question isn't clear enough.
As part of my master's thesis, I have to plot bounded regions of the night sky onto a 2D plane. My current solution consists of a rectangular mapping where (ra, dec) values are plotted to (x,y) coordinates. While this approach works well enough for small regions in relatively low ascension values, the resulting plots get progressively distorted for higher ||dec|| values, as expected.
At some point I'll have to change this to a more versatile approach. Thing is, I'm not exactly clear on what to search for. I guess I have to be able to map angular coordinates to a square (or hexagon) subgrid, but most search results I get are concerned with full-surface mapping.
I know I won't be able to achieve a perfect, distortion-free plotting, but I don't require perfect solutions; only a more general projection that will work well near the poles. Something like this, where I put my Photoshop skills to work and try to simulate a 20ยบ region under my current approach and the one I'm looking for:
What I want:
What I have:
TL;DR: how do I convert between coordinates on a sphere (ra/dec) to cartesian coordinates on a locally-defined grid?
Let me explain my problem:
I have a black vector shape (let's say it's a series of joined, straight lines for now, but it'd be nice if I could also support quadratic curves).
I also have a rectangle of a predefined width and height. I'm going to place it on top of the black shape, and then take the union of the two.
My first issue is that I don't know how to quickly extract vector unions, but I think there is a well-defined formula I can figure out for myself.
My second, and more tricky issue is how to efficiently detect the position the rectangle needs to be in (i.e., what translation and rotation are needed by the matrices), in order to maximize the black, remaining after the union (see figure, below).
The red outlined shape below is ~33% black; the green is something like 85%; and there are positions for this shape & rectangle wherein either could have 100% coverage.
Obviously, I can brute-force this by trying every translation and rotation value for every point where at least part of the rectangle is touching the black shape, then keep track of the one with the most black coverage. The problem is, I can only try a finite number of positions (and may therefore miss the maximum). Apart from that, it feels very inefficient!
Can you think of a more efficient way of tackling this problem?
Something from my Uni days tells me that a Fourier transform might improve the efficiency here, but I can't figure out how I'd do that with a vector shape!
Three ideas that have promise of being faster and/or more precise than brute force search:
Suppose you have a 3d physics engine. Define a "cone-shaped" surface where the apex is at say (0,0,-1), the black polygon boundary on the z=0 plane with its centroid at the origin, and the cone surface is formed by connecting the apex with semi-infinite rays through the polygon boundary. Think of a party hat turned upside down and crumpled to the shape of the black polygon. Now constrain the rectangle to be parallel to the z=0 plane and initially so high above the cone (large z value) that it's easy to find a place where it's definitely "inside". Then let the rectangle fall downward under gravity, twisting about z and translating in x-y only as it touches the cone, staying inside all the way down until it settles and can't move any farther. The collision detection and force resolution of the physics engine takes care of the complexities. When it settles, it will be in a position of maximal coverage of the black polygon in a local sense. (If it settles with z<0, then coverage is 100%.) For the convex case it's probably a global maximum. To probabilistically improve the result for non-convex cases (like your example), you'd randomize the starting position, dropping the polygon many times, taking the best result. Note you don't really need a full blown physics engine (though they certainly exist in open source). It's enough to use collision resolution to tell you how to rotate and translate the rectangle in a pseudo-physical way as it twists and slides uniformly down the z axis as far as possible.
Different physics model. Suppose the black area is an attractive field generator in 2d following the usual inverse square rule like gravity and magnetism. Now let the rectangle drift in a damping medium responding to this field. It ought to settle with a maximal area overlapping the black area. There are problems with "nulls" like at the center of a donut, but I don't think these can ever be stable equillibria. Can they? The simulation could be easily done by modeling both shapes as particle swarms. Or since the rectangle is a simple shape and you are a physicist, you could come up with a closed form for the integral of attractive force between a point and the rectangle. This way only the black shape needs representation as particles. Come to think of it, if you can come up with a closed form for torque and linear attraction due to two triangles, then you can decompose both shapes with a (e.g. Delaunay) triangulation and get a precise answer. Unfortunately this discussion implies it can't be done analytically. So particle clouds may be the final solution. The good news is that modern processors, particularly GPUs, do very large particle computations with amazing speed. Edit: I implemented this quick and dirty. It works great for convex shapes, but concavities create stable points that aren't what you want. Using the example:
This problem is related to robot path planning. Looking at this literature may turn up some ideas In RPP you have obstacles and a robot and want to find a path the robot can travel while avoiding and/or sliding along them. If the robot is asymmetric and can rotate, then 2d planning is done in a 3d (toroidal) configuration space (C-space) where one dimension is rotation (so closes on itself). The idea is to "grow" the obstacles in C-space while shrinking the robot to a point. Growing the obstacles is achieved by computing Minkowski Differences.) If you decompose all polygons to convex shapes, then there is a simple "edge merge" algorithm for computing the MD.) When the C-space representation is complete, any 1d path that does not pierce the "grown" obstacles corresponds to continuous translation/rotation of the robot in world space that avoids the original obstacles. For your problem the white area is the obstacle and the rectangle is the robot. You're looking for any open point at all. This would correspond to 100% coverage. For the less than 100% case, the C-space would have to be a function on 3d that reflects how "bad" the intersection of the robot is with the obstacle rather than just a binary value. You're looking for the least bad point. C-space representation is an open research topic. An octree might work here.
Lots of details to think through in both cases, and they may not pan out at all, but at least these are frameworks to think more about the problem. The physics idea is a bit like using simulated spring systems to do graph layout, which has been very successful.
I don't believe it is possible to find the precise maximum for this problem, so you will need to make do with an approximation.
You could potentially render the vector image into a bitmap and use Haar features for this - they provide a very quick O(1) way of calculating the average colour of a rectangular region.
You'd still need to perform this multiple times for different rotations and positions, but it would bring it algorithmic complexity down from a naive O(n^5) to O(n^3) which may be acceptably fast. (with n here being the size of the different degrees of freedom you are scanning)
Have you thought to keep track of the remaining white space inside the blocks with something like if whitespace !== 0?
Am following this tutorial for my 2d game collision handling , this tutorial explains about the collision used in one of my favorite game "N". How they used separate axis theorem more efficiently for collision between AABB vs AABB and AABB vs Circle. http://www.metanetsoftware.com/technique/tutorialA.html. I understand the implementation of AABB vs AABB collision handling but I couldn't understand AABB vs Circle collision detection especially voronoi regions.Totally confused how/where to start.
AABB vs AABB collision detection
Find the axis along all the edge by finding the normal of each edge.
Projection all the vertices to the
resultant Axis , final result should
be a scalar value.
The resultant scalar value in turn
is used to find whether collision is
present or not.
Can someone please explain how to handle collision AABB vc Circle - vise versa?
Since collisions with a circle always come down to a comparison against the radius (in your case, via projection), having the closest line segment (edge of the polygon) and the normal vector are the only building blocks you need. The normal vector is easily computed from the points of the line segment (something like unit(y2-y1, x1-x2) ... the negative reciprocal of the slope). Figuring out which edge is closest is the building block that remains. Voronoi regions give us the last building block.
You understand collisions between axis-aligned bounding boxes. I assume you also understand collisions between two circles. I'm assuming you don't understand voronoi regions. So, where to start? Voronoi diagrams. I highly suggest that you find a diagrammed explanation. This link is quite good. However, depending on how lost you are, perhaps a little additional background (seriously, though, no explanation can beat the visual):
A voronoi diagram is one of the ubiquitous data structures of computational geometry. Any computational geometry book will discuss the Voronoi diagram. It answers a simple question: where is the closest post office? Given a set of points in a plane (post offices), a voronoi diagram separates the plane into different regions, each containing one of the points. If you are in a particular region, you know which point (post office) is closest to you. If you were a circle, this would be nice for collision detection for a simple reason: the closest point is the most important one to test for collisions.
Note that if you want to mathematically derive a voronoi diagram, you simply consider all point pairs and calculate all bisecting lines. Then you intersect all of the bisecting lines and throw away the segments that are unimportant because some other point is closer to the point of interest (which happens at every intersection). This leads to a terribly inefficient algorithm, though. The efficient implementation involves another ubiquitous thing in computational geometry: the line-sweep algorithm. Its details can be found elsewhere; the important bit is that it provides a method of considering only the important points at any stage of the algorithm.
The voronoi regions in your tutorial are a little more complex. Instead of just points, we have line segments. Fortunately, the line-sweep algorithm handles this nicely. You mostly have to worry about the start or end of the line segments. Conceptually, not much changes once you have the basic algorithm down. Again, this is exceptionally helpful for collision detection with a circle: given the voronoi region, you know which line segment to test collisions against.
Does that even help? Feedback appreciated. I'll be happy to clarify anything. Explaining voronoi diagrams without visuals is probably a bad idea.
I have a 2D computational geometry / GIS problem that I think should be common and I'm hoping to find some existing code/library to use.
The problem is to check which subset of a big set (thousands) of small polygons intersect with a single large polygon. (By "small" and "large" I'm referring to the amount of space the polygons cover, not the number of points that define them, although in general suppose that the number of points defining a polygon is roughly proportional to its geometric size. And to give a sense of proportion, think of "large" as the polygon for a state in the United States, and "small" as the polygon for a town.)
Suppose the naive solution using a standard CheckIfPolygonsIntersect( P, p ) function, called for each small polygon p against the one large polygon P, is too slow. It seems that there are ways to pre-process the large polygon to make the intersection checks for the majority of the small polygons trivial. In particular, it seems like you could create a small set of rectangles that partially/almost fill the large polygon. And similarly you could create a small set of rectangles that partially/almost fill the area of the bounding box of the large polygon that is not actually within the large polygon. Then the vast majority of your small polygons could be trivially included or excluded: if they are fully outside the bounding rect of the large polygon, they are excluded. If they are fully inside the boundary of one of the inside-bounding-rect-but-outside-polygon rects, they are excluded. If any of their points are within any of the internal rects, they are included. And only if none of the above apply do you have to call the CheckIfPolygonsIntersect( P, p ) function.
Is that a well-known algorithm? Do you know of existing code to compute a reasonable set of interior/exterior rectangles for arbitrary (convex or concave) polygons? The rectangles don't have to be perfect in all cases; they just have to fill much of the polygon, and much of the inside-bounding-rect-but-outside-polygon area.
Here's a simple plan for how I might compute these rectangles:
take the bounding box of the large polygon and construct a, say, 10x10 grid of points over it
for each point, determine if it's inside or outside the polygon
"grow" each point into a rectangle by iteratively expanding it in each of the four directions until one of the rect edges crosses one of the polygon edges, in which case you've gone too far (this would actually be done in a "binary search" kind of iteration so with just a few iterations you could find the correct amount to expand in each direction; and of course there is some question of whether to maximize the edges one at a time or in concert with one another)
any not-yet-expanded grid point that get covered by another point's expansion just disappears
when all points have been expanded (or have disappeared), you have your set of interior and exterior rectangles
Of course, certain crazy concave shapes for the large polygon could lead to some poor/small rectangles. But assuming the polygons are mostly reasonable (e.g., say they were the shapes of the states of the United States), it seems like you'd get a good set of rectangles and could greatly optimize those thousands of intersection checks you'd subsequently do.
Is there a name (and code) for that algorithm?
Edit: I am already using a quad-tree to determine the small polygons that are likely to intersect with the bounding rect of the large polygon. So the problem is about checking which of those polygons actually do intersect with the large polygon.
Thanks for any help.
In your plan you described something very similar to the signed distance map method. Google 'distance map algorithm' for details. I hope it will be what you're looking for.
I have an array of points. I want to know if this array of point represents a circle, a square or a triangle.
Where should i begin? (i use C#)
Thanks
Jon
Depending on your problem, a good approach for this problem may be to use the Hough transform and all its derived algorithm
It consists in a transformation of the image space to an other space where the coordinate represents the objects parameters (angle and initial point for a line, coordinates of the center and radius for a circle)
The algorithm transforms each point of your array of points in points in the other space. Then you have to search in the new space if some points are prevailing. From these points, you will get the parameters of your object.
Of course, you need to do it once to recognize the lines (so you will know how many lines are in your bitmap and where they are) and to it to recognize the circles (it is not exactly the same algorithm)
You may have a look to this lecture (for Hough Circle Transform), but you could easily find the algorithm for line
EDIT: you can also have a look to these answers
Shape recognition algorithm(s)
Detecting an object on the image based on geometrical form
imagine it is each of these one-by-one and try to fit each of these shapes on the data.. for a square, you could find the four extreme points, and try charting out a square that goes through all of them..
Once you have got a shape in place.. you could measure the distance between each of the points and the part of the shape that is nearest to it.. then square these distances and add them up.. the shape which has the smallest sum-of-squares is probably your best bet
Use the Hough Transform.
I'm going to take a wild stab and say if you have 3 points the shape represents a triangle, 4 points is some kind of quadrilateral, any more than that is a circle.
Perhaps there's more information to your problem you could provide.