Related
Given a "density" scalar field in the plane, how can I divide the plane into nice (low moment of inertia) regions so that each region contains a similar amount of "mass"?
That's not the best description of what my actual problem is, but it's the most concise phrasing I could think of.
I have a large map of a fictional world for use in a game. I have a pretty good idea of approximately how far one could walk in a day from any given point on this map, and this varies greatly based on the terrain etc. I would like to represent this information by dividing the map into regions, so that one day of walking could take you from any region to any of its neighboring regions. It doesn't have to be perfect, but it should be significantly better than simply dividing the map into a hexagonal grid (which is what many games do).
I had the idea that I could create a gray-scale image with the same dimensions as the map, where each pixel's color value represents how quickly one can travel through the pixel in the same place on the map. Well-maintained roads would be encoded as white pixels, and insurmountable cliffs would be encoded as black, or something like that.
My question is this: does anyone have an idea of how to use such a gray-scale image (the "density" scalar field) to generate my "grid" from the previous paragraph (regions of similar "mass")?
I've thought about using the gray-scale image as a discrete probability distribution, from which I can generate a bunch of coordinates, and then use some sort of clustering algorithm to create the regions, but a) the clustering algorithms would have to create clusters of a similar size, I think, for that idea to work, which I don't think they usually do, and b) I barely have any idea if any of that even makes sense, as I'm way out of my comfort zone here.
Sorry if this doesn't belong here, my idea has always been to solve it programatically somehow, so this seemed the most sensible place to ask.
UPDATE: Just thought I'd share the results I've gotten so far, trying out the second approach suggested by #samgak - recursively subdividing regions into boxes of similar mass, finding the center of mass of each region, and creating a voronoi diagram from those.
I'll keep tweaking, and maybe try to find a way to make it less grid-like (like in the upper right corner), but this worked way better than I expected!
Building upon #samgak's solution, if you don't want the grid-like structure, you can just add a small random perturbation to your centers. You can see below for example the difference I obtain:
without perturbation
adding some random perturbation
A couple of rough ideas:
You might be able to repurpose a color-quantization algorithm, which partitions color-space into regions with roughly the same number of pixels in them. You would have to do some kind of funny mapping where the darker the pixel in your map, the greater the number of pixels of a color corresponding to that pixel's location you create in a temporary image. Then you quantize that image into x number of colors and use their color values as co-ordinates for the centers of the regions in your map, and you could then create a voronoi diagram from these points to define your region boundaries.
Another approach (which is similar to how some color quantization algorithms work under the hood anyway) could be to recursively subdivide regions of your map into axis-aligned boxes by taking each rectangular region and choosing the optimal splitting line (x or y) and position to create 2 smaller rectangles of similar "mass". You would end up with a power of 2 count of rectangular regions, and you could get rid of the blockiness by taking the centre of mass of each rectangle (not simply the center of the bounding box) and creating a voronoi diagram from all the centre-points. This isn't guaranteed to create regions of exactly equal mass, but they should be roughly equal. The algorithm could be improved by allowing recursive splitting along lines of arbitrary orientation (or maybe a finite number of 8, 16, 32 etc possible orientations) but of course that makes it more complicated.
The scenario : There is a rectangular space inside which there are arbitrarily placed polygons of arbitrary orientations. The aim is to find the largest empty rectangle that can be fitted inside the empty regions of the rectangular space. These images below illustrate the scenario with the polygons in blue and the dotted line representing the maximum empty rectangle that can be fitted in each scenario.
The problem : Apparently, finding largest empty rectangles is a well known problem in computational geometry, but the algorithms I found in this area dealt with finding empty rectangles amid points (CGAL has implemented this) and line segments. Is there a way to adapt these existing techniques for my scenario? Or is there a simpler way to do this?
Unfortunately, most of the computational geometry literature with which I am familiar seems to generate beautiful descriptions of algorithms and proofs of their correctness without actually providing implementations. Perhaps this is because the implementations are generally rather involved.
You don't mention what degree of inaccuracy you can tolerate. If you have some tolerance, this answer's for you.
My suggestion is that you turn this hard problem into an easier problem.
Find the bounding box of your polygon collection.
Divide the bounding box into a grid. The finer the grid the better your accuracy, but the longer it will take to find a solution.
Find how much area of each grid cell (cast as a rectangular polygon) intersects with the polygon set.
If the overlap is sufficient (greater than some minimum value you specify), mark the grid cell with a zero; otherwise, mark it with a one.
You now have a rectangular array of zeros and ones. This forms the basis of the easier problem: what is the largest rectangular subset of this grid which is composed entirely of ones?
This easier problem has a number of accessible solutions all over the internet (e.g. 1, 2, 3, 4, 5, 6).
I was wondering if anybody could point me to the best algorithm/heuristic which will fit my particular polygon packing problem. I am given a single polygon as a boundary (convex or concave may also contain holes) and a single "fill" polygon (may also be convex or concave, does not contain holes) and I need to fill the boundary polygon with a specified number of fill polygons. (I'm working in 2D).
Many of the polygon packing heuristics I've found assume that the boundary and/or filling polygons will be rectangular and also that the filling polygons will be of different sizes. In my case, the filling polygons may be non-rectangular, but all will be exactly the same.
Maybe this is a particular type of packing problem? If somebody has a definition for this type of polygon packing I'll gladly google away, but so far I've not found anything which is similar enough to be of great use.
Thanks.
The question you ask is very hard. To put this in perspective, the (much) simpler case where you're packing the interior of your bounded polygon with non-overlapping disks is already hard, and disks are the simplest possible "packing shape" (with any other shape you have to consider orientation as well as size and center location).
In fact, I think it's an open problem in computational geometry to determine for an arbitrary integer N and arbitrary bounded polygonal region (in the Euclidean plane), what is the "optimal" (in the sense of covering the greatest percentage of the polygon interior) packing of N inscribed non-overlapping disks, where you are free to choose the radius and center location of each disk. I'm sure the "best" answer is known for certain special polygonal shapes (like rectangles, circles, and triangles), but for arbitrary shapes your best "heuristic" is probably:
Start your shape counter at N.
Add the largest "packing shape" you can fit completely inside the polygonal boundary without overlapping any other packing shapes.
Decrement your shape counter.
If your shape counter is > 0, go to step 2.
I say "probably" because "largest first" isn't always the best way to pack things into a confined space. You can dig into that particular flavor of craziness by reading about the bin packing problem and knapsack problem.
EDIT: Step 2 by itself is hard. A reasonable strategy would be to pick an arbitrary point on the interior of the polygon as the center and "inflate" the disk until it touches either the boundary or another disk (or both), and then "slide" the disk while continuing to inflate it so that it remains inside the boundary without overlapping any other disks until it is "trapped" - with at least 2 points of contact with the boundary and/or other disks. But it isn't easy to formalize this "sliding process". And even if you get the sliding process right, this strategy doesn't guarantee that you'll find the biggest "inscribable disk" - your "locally maximal" disk could be trapped in a "lobe" of the interior which is connected by a narrow "neck" of free space to a larger "lobe" where a larger disk would fit.
Thanks for the replies, my requirements were such that I was able to further simplify the problem by not having to deal with orientation and I then even further simplified by only really worrying about the bounding box of the fill element. With these two simplifications the problem became much easier and I used a stripe like filling algorithm in conjunction with a spatial hash grid (since there were existing elements I was not allowed to fill over).
With this approach I simply divided the fill area into stripes and created a spatial hash grid to register existing elements within the fill area. I created a second spatial hash grid to register the fill area (since my stripes were not guaranteed to be within the bounding area, this made checking if my fill element was in the fill area a little faster since I could just query the grid and if all grids where my fill element were to be placed, were full, I knew the fill element was inside the fill area). After that, I iterated over each stripe and placed a fill element where the hash grids would allow. This is certainly not an optimal solution, but it ended up being all that was required for my particular situation and pretty fast as well. I found the required information about creating a spatial hash grid from here. I got the idea for filling by stripes from this article.
This type of problem is very complex to solve geometrically.
If you can accept a good solution instead of the 100% optimal
solution then you can to solve it with a raster algorithm.
You draw (rasterize) the boundary polygon into one in-memory
image and the fill polygon into another in-memory image.
You can then more easily search for a place where the fill polygon will
fit in the boundary polygon by overlaying the two images with
various (X, Y) offsets for the fill polygon and checking
the pixel values.
When you find a place that the fill polygon fits,
you clear the pixels in the boundary polygon and repeat
until there are no more places where the fill polygon fits.
The keywords to google search for are: rasterization, overlay, algorithm
If your fill polygon is the shape of a jigsaw piece, many algorithms will miss the interlocking alignment. (I don't know what to suggest in that case)
One approach to the general problem that works well when the boundary is much larger than
the fill pieces is to tile an infinite plane with the pieces in the best way you can, and then look for the optimum alignment of the boundary on this plane.
I have a map that is cut up into a number of regions by borders (contours) like countries on a world map. Each region has a certain surface-cover class S (e.g. 0 for water, 0.03 for grass...). The borders are defined by:
what value of S is on either side of it (0.03 on one side, 0.0 on the other, in the example below)
how many points the border is made of (n=7 in example below), and
n coordinate pairs (x, y).
This is one example.
0.0300 0.0000 7
2660607.5 6332685.5 2660565.0 6332690.5 2660541.5 6332794.5
2660621.7 6332860.5 2660673.8 6332770.5 2660669.0 6332709.5
2660607.5 6332685.5
I want to make a raster map in which each pixel has the value of S corresponding to the region in which the center of the pixel falls.
Note that the borders represent step changes in S. The various values of S represent discrete classes (e.g. grass or water), and are not values that can be averaged (i.e. no wet grass!).
Also note that not all borders are closed loops like the example above. This is a bit like country borders: e.g. the US-Canada border isn't a closed loop, but rather a line joining up at each end with two other borders: the Canada-ocean and the US-ocean "borders". (Closed-loop borders do exist nevertheless!)
Can anyone point me to an algorithm that can do this? I don't want to reinvent the wheel!
The general case for processing this sort of geometry in vector form can be quite difficult, especially since nothing about the structure you describe requires the geometry to be consistent. However, since you just want to rasterize it, then treating the problem as a Voronoi diagram of line segments can be more robust.
Approximating the Voronoi diagram can be done graphically in OpenGL by drawing each line segment as a pair of quads making a tent shape. The z-buffer is used to make the closest quad take precedence, and thus color the pixel based on whichever line is closest. The difference here is that you will want to color the polygons based on which side of the line they are on, instead of which line they represent. A good paper discussing a similar algorithm is Hoff et al's Fast Computation of Generalized Voronoi Diagrams Using Graphics Hardware
The 3d geometry will look something like this sketch with 3 red/yellow segments and 1 blue/green segment:
This procedure doesn't require you to convert anything into a closed loop, and doesn't require any fancy geometry libraries. Everything is handled by the z-buffer, and should be fast enough to run in real time on any modern graphics card. A refinement would be to use homogeneous coordinates to make the bases project to infinity.
I implemented this algorithm in a Python script at http://www.pasteall.org/9062/python. One interesting caveat is that using cones to cap the ends of the lines didn't work without distorting the shape of the cone, because the cones representing the end points of the segments were z-fighting. For the sample geometry you provided, the output looks like this:
I'd recommend you to use a geometry algorithm library like CGAL. Especially the second example in the "2D Polygons" page of the reference manual should provide you what you need. You can define each "border" as a polygon and check if certain points are inside the polygons. So basically it would be something like
for every y in raster grid
for every x in raster grid
for each defined polygon p
if point(x,y) is inside polygon p
pixel[X][Y] = inside_color[p]
I'm not so sure about what to do with the outside_color because the outside regions will overlap, won't they? Anyway, looking at your example, every outside region could be water, so you just could do a final
if pixel[X][Y] still undefined then pixel[X][Y] = water_value
(or as an alternative, set pixel[X][Y] to water_value before iterating through the polygon list)
first, convert all your borders into closed loops (possibly including the edges of your map), and indentify the inside colour. this has to be possible, otherwise you have an inconsistency in your data
use bresenham's algorithm to draw all the border lines on your map, in a single unused colour
store a list of all the "border pixels" as you do this
then for each border
triangulate it (delaunay)
iterate through the triangles till you find one whose centre is inside your border (point-in-polygon test)
floodfill your map at that point in the border's interior colour
once you have filled in all the interior regions, iterate through the list of border pixels, seeing which colour each one should be
choose two unused colors as markers "empty" and "border"
fill all area with "empty" color
draw all region borders by "border" color
iterate through points to find first one with "empty" color
determine which region it belongs to (google "point inside polygon", probably you will need to make your borders closed as Martin DeMello suggested)
perform flood-fill algorithm from this point with color of the region
go to next "empty" point (no need to restart search - just continue)
and so on till no "empty" points will remain
The way I've solved this is as follows:
March along each segment; stop at regular intervals L.
At each stop, place a tracer point immediately to the left and to the right of the segment (at a certain small distance d from the segment). The tracer points are attributed the left and right S-value, respectively.
Do a nearest-neighbour interpolation. Each point on the raster grid is attributed the S of the nearest tracer point.
This works even when there are non-closed lines, e.g. at the edge of the map.
This is not a "perfect" analytical algorithm. There are two parameters: L and d. The algorithm works beautifully as long as d << L. Otherwise you can get inaccuracies (usually single-pixel) near segment junctions, especially those with acute angles.
I've been working on a visualization project for 2-dimensional continuous data. It's the kind of thing you could use to study elevation data or temperature patterns on a 2D map. At its core, it's really a way of flattening 3-dimensions into two-dimensions-plus-color. In my particular field of study, I'm not actually working with geographical elevation data, but it's a good metaphor, so I'll stick with it throughout this post.
Anyhow, at this point, I have a "continuous color" renderer that I'm very pleased with:
The gradient is the standard color-wheel, where red pixels indicate coordinates with high values, and violet pixels indicate low values.
The underlying data structure uses some very clever (if I do say so myself) interpolation algorithms to enable arbitrarily deep zooming into the details of the map.
At this point, I want to draw some topographical contour lines (using quadratic bezier curves), but I haven't been able to find any good literature describing efficient algorithms for finding those curves.
To give you an idea for what I'm thinking about, here's a poor-man's implementation (where the renderer just uses a black RGB value whenever it encounters a pixel that intersects a contour line):
There are several problems with this approach, though:
Areas of the graph with a steeper slope result in thinner (and often broken) topo lines. Ideally, all topo lines should be continuous.
Areas of the graph with a flatter slope result in wider topo lines (and often entire regions of blackness, especially at the outer perimeter of the rendering region).
So I'm looking at a vector-drawing approach for getting those nice, perfect 1-pixel-thick curves. The basic structure of the algorithm will have to include these steps:
At each discrete elevation where I want to draw a topo line, find a set of coordinates where the elevation at that coordinate is extremely close (given an arbitrary epsilon value) to the desired elevation.
Eliminate redundant points. For example, if three points are in a perfectly-straight line, then the center point is redundant, since it can be eliminated without changing the shape of the curve. Likewise, with bezier curves, it is often possible to eliminate cetain anchor points by adjusting the position of adjacent control points.
Assemble the remaining points into a sequence, such that each segment between two points approximates an elevation-neutral trajectory, and such that no two line segments ever cross paths. Each point-sequence must either create a closed polygon, or must intersect the bounding box of the rendering region.
For each vertex, find a pair of control points such that the resultant curve exhibits a minimum error, with respect to the redundant points eliminated in step #2.
Ensure that all features of the topography visible at the current rendering scale are represented by appropriate topo lines. For example, if the data contains a spike with high altitude, but with extremely small diameter, the topo lines should still be drawn. Vertical features should only be ignored if their feature diameter is smaller than the overall rendering granularity of the image.
But even under those constraints, I can still think of several different heuristics for finding the lines:
Find the high-point within the rendering bounding-box. From that high point, travel downhill along several different trajectories. Any time the traversal line crossest an elevation threshold, add that point to an elevation-specific bucket. When the traversal path reaches a local minimum, change course and travel uphill.
Perform a high-resolution traversal along the rectangular bounding-box of the rendering region. At each elevation threshold (and at inflection points, wherever the slope reverses direction), add those points to an elevation-specific bucket. After finishing the boundary traversal, start tracing inward from the boundary points in those buckets.
Scan the entire rendering region, taking an elevation measurement at a sparse regular interval. For each measurement, use it's proximity to an elevation threshold as a mechanism to decide whether or not to take an interpolated measurement of its neighbors. Using this technique would provide better guarantees of coverage across the whole rendering region, but it'd be difficult to assemble the resultant points into a sensible order for constructing paths.
So, those are some of my thoughts...
Before diving deep into an implementation, I wanted to see whether anyone else on StackOverflow has experience with this sort of problem and could provide pointers for an accurate and efficient implementation.
Edit:
I'm especially interested in the "Gradient" suggestion made by ellisbben. And my core data structure (ignoring some of the optimizing interpolation shortcuts) can be represented as the summation of a set of 2D gaussian functions, which is totally differentiable.
I suppose I'll need a data structure to represent a three-dimensional slope, and a function for calculating that slope vector for at arbitrary point. Off the top of my head, I don't know how to do that (though it seems like it ought to be easy), but if you have a link explaining the math, I'd be much obliged!
UPDATE:
Thanks to the excellent contributions by ellisbben and Azim, I can now calculate the contour angle for any arbitrary point in the field. Drawing the real topo lines will follow shortly!
Here are updated renderings, with and without the ghetto raster-based topo-renderer that I've been using. Each image includes a thousand random sample points, represented by red dots. The angle-of-contour at that point is represented by a white line. In certain cases, no slope could be measured at the given point (based on the granularity of interpolation), so the red dot occurs without a corresponding angle-of-contour line.
Enjoy!
(NOTE: These renderings use a different surface topography than the previous renderings -- since I randomly generate the data structures on each iteration, while I'm prototyping -- but the core rendering method is the same, so I'm sure you get the idea.)
Here's a fun fact: over on the right-hand-side of these renderings, you'll see a bunch of weird contour lines at perfect horizontal and vertical angles. These are artifacts of the interpolation process, which uses a grid of interpolators to reduce the number of computations (by about 500%) necessary to perform the core rendering operations. All of those weird contour lines occur on the boundary between two interpolator grid cells.
Luckily, those artifacts don't actually matter. Although the artifacts are detectable during slope calculation, the final renderer won't notice them, since it operates at a different bit depth.
UPDATE AGAIN:
Aaaaaaaand, as one final indulgence before I go to sleep, here's another pair of renderings, one in the old-school "continuous color" style, and one with 20,000 gradient samples. In this set of renderings, I've eliminated the red dot for point-samples, since it unnecessarily clutters the image.
Here, you can really see those interpolation artifacts that I referred to earlier, thanks to the grid-structure of the interpolator collection. I should emphasize that those artifacts will be completely invisible on the final contour rendering (since the difference in magnitude between any two adjacent interpolator cells is less than the bit depth of the rendered image).
Bon appetit!!
The gradient is a mathematical operator that may help you.
If you can turn your interpolation into a differentiable function, the gradient of the height will always point in the direction of steepest ascent. All curves of equal height are perpendicular to the gradient of height evaluated at that point.
Your idea about starting from the highest point is sensible, but might miss features if there is more than one local maximum.
I'd suggest
pick height values at which you will draw lines
create a bunch of points on a fine, regularly spaced grid, then walk each point in small steps in the gradient direction towards the nearest height at which you want to draw a line
create curves by stepping each point perpendicular to the gradient; eliminate excess points by killing a point when another curve comes too close to it-- but to avoid destroying the center of hourglass like figures, you might need to check the angle between the oriented vector perpendicular to the gradient for both of the points. (When I say oriented, I mean make sure that the angle between the gradient and the perpendicular value you calculate is always 90 degrees in the same direction.)
In response to your comment to #erickson and to answer the point about calculating the gradient of your function. Instead of calculating the derivatives of your 300 term function you could do a numeric differentiation as follows.
Given a point [x,y] in your image you could calculate the gradient (direction of steepest decent)
g={ ( f(x+dx,y)-f(x-dx,y) )/(2*dx),
{ ( f(x,y+dy)-f(x,y-dy) )/(2*dy)
where dx and dy could be the spacing in your grid. The contour line will run perpendicular to the gradient. So, to get the contour direction, c, we can multiply g=[v,w] by matrix, A=[0 -1, 1 0] giving
c = [-w,v]
Alternately, there is the marching squares algorithm which seems appropriate to your problem, although you may want to smooth the results if you use a coarse grid.
The topo curves you want to draw are isosurfaces of a scalar field over 2 dimensions. For isosurfaces in 3 dimensions, there is the marching cubes algorithm.
I've wanted something like this myself, but haven't found a vector-based solution.
A raster-based solution isn't that bad, though, especially if your data is raster-based. If your data is vector-based too (in other words, you have a 3D model of your surface), you should be able to do some real math to find the intersection curves with horizontal planes at varying elevations.
For a raster-based approach, I look at each pair of neighboring pixels. If one is above a contour level, and one is below, obviously a contour line runs between them. The trick I used to anti-alias the contour line is to mix the contour line color into both pixels, proportional to their closeness to the idealized contour line.
Maybe some examples will help. Suppose that the current pixel is at an "elevation" of 12 ft, a neighbor is at an elevation of 8 ft, and contour lines are every 10 ft. Then, there is a contour line half way between; paint the current pixel with the contour line color at 50% opacity. Another pixel is at 11 feet and has a neighbor at 6 feet. Color the current pixel at 80% opacity.
alpha = (contour - neighbor) / (current - neighbor)
Unfortunately, I don't have the code handy, and there might have been a bit more to it (I vaguely recall looking at diagonal neighbors too, and adjusting by sqrt(2) / 2). I hope this enough to give you the gist.
It occurred to me that what you're trying to do would be pretty easy to do in MATLAB, using the contour function. Doing things like making low-density approximations to your contours can probably be done with some fairly simple post-processing of the contours.
Fortunately, GNU Octave, a MATLAB clone, has implementations of the various contour plotting functions. You could look at that code for an algorithm and implementation that's almost certainly mathematically sound. Or, you might just be able to offload the processing to Octave. Check out the page on interfacing with other languages to see if that would be easier.
Disclosure: I haven't used Octave very much, and I haven't actually tested it's contour plotting. However, from my experience with MATLAB, I can say that it will give you almost everything you're asking for in just a few lines of code, provided you get your data into MATLAB.
Also, congratulations on making a very VanGough-esque slopefield plot.
I always check places like http://mathworld.wolfram.com before going to deep on my own :)
Maybe their curves section would help? Or maybe the entry on maps.
compare what you have rendered with a real-world topo map - they look identical to me! i wouldn't change a thing...
Write the data out as an HGT file (very simple digital elevation data format used by USGS) and use the free and open-source gdal_contour tool to create contours. That works very well for terrestrial maps, the constraint being that the data points are signed 16-bit numbers, which fits the earthly range of heights in metres very well, but may not be enough for your data, which I assume not to be a map of actual terrain - although you do mention terrain maps.
I recommend the CONREC approach:
Create an empty line segment list
Split your data into regular grid squares
For each grid square, split the square into 4 component triangles:
For each triangle, handle the cases (a through j):
If a line segment crosses one of the cases:
Calculate its endpoints
Store the line segment in the list
Draw each line segment in the line segment list
If the lines are too jagged, use a smaller grid. If the lines are smooth enough and the algorithm is taking too long, use a larger grid.