Minkowski Sum of 3D(+) convex polygons - python

My goal is to obtain the representations of all faces (in the form of A[x,y,z]'>b) of a polyhedron that is the result of the convex difference between two convex polyhedra. Meaning, finding the intersection of all planes that are the result of the Minkowski difference of P1 - P2 = { x - y | x \in P1, y \in P2 }.
I'm looking for either an established library (Python?) or an idea on how to do this efficiently. I thought about doing something similar to the GJK algorithm but I need all of the faces, and not just compute whether the origin is inside quickly. Moreover, seems inefficient to use this support function in a methodological way in 3D, or higher dimensions. Also, let's say I got the vertices, do I now need to form the plane equation from two vectors on it with the cross product, for every face, or is there a way to obtain it from the Minkowski sum itself? (keeping in mind the need for higher dimensions).

Ok, it seems I was finally able to solve it, and I'm posting in case it would interest anyone in the future:
First, I pip installed the pypoman library.
With it, we are able to move easily between vertices and faces with compute_polytope_halfspaces (aka, the H-representation of a polytope). So I get the representation P_i: H_i x < h_i for i=1,2 from the vertices (or skip it if it's already in the correct format).
Now if we set P_sum = {[x1;x2] \in R^2n | [H_1 0; 0 H_2] [x1;x2]' < [h_1,h_2]'}, notice that the Minkowski sum is equivalent to P1+P2 = [I,I] P_sum (idea from this paper IV.B). So I can use pypoman's project_polytope function to get the Minkwoski sum with H_sum x < h_sum in the original dimensions.

Related

I want to calculate overlap area from semicircle draw in map using python

I'm trying calculate overlap area created by semicircle which created by folium.after that i want to assign x,y,z value to the each semi semicircle by reduce same value overlap ex:- to reduce two x value semicircles
here is the image for better understand
How do I calculate the overlap area using python?
1. Pure Math:
First, check if they intersect:
Since the overlaps are not uniform (orthogonal, perpendicular or symmetrical) you are going to need to use the help of something called FUNCTIONS (like f(x)) and since your sectors (they are not a semi-circle) look to be the same size and have a internal area, you can use INEQUALITIES to check if they are intersecting.
Say for example you have a two sectors that are intercepting like this, with their respective functions:
f(x) = x^2 + y^2 <= 5 | {0 <= x <= 5} {0 <= y} (RED)
g(x) = ((x - 4)^2) + y^2 <= 5 | {0 <= x <= 5} {0 <= y} (BLUE)
Assuming the unanimous cartesian system (x, y), all the sectors are drawing to the same coordinates (in the same 'space'; if you understand me here...):
Just equate the 2 functions together:
Rule: If f(x) intersects g(x) and vice-versa:
f(x) = g(x) at some point
g(x) is in / passes through the shaded area of f(x)
f(x) is in / passes through the shaded area of g(x)
The radius of f(x) + the radius of g(x) ≥ the distance between the 2 sectors
Note: 2 & 3 might need calculus to check, if your functions are not inequalities but lines.
Since you are coding a program, you can do an early out, that if one of the top 4 conditions are not met, the sectors do not intersect.
Second, if the sectors do intersect:
Use intergral calculus to find the area of intersection:
∫f(x) - ∫g(x) = the magnitude (size) of the area (make sure the upper and lower limits are specified correctly)
2. Using a 3D graphics method:
Now, I know this is not a really good method for this, but you could implement collision detection, if you do not like heavy maths.
Obviously, at first reading this may sound completely absurd (because collision detection is something used in game programming far more often), but have read at this:
circle-circle collision
and see if you can possibly implement a sort of sector hit-box on your quadrants (create circle hit-boxes, restrict them to sectors, and see if they intersect.)
Honestly, though, the first method would be the most optimum, assuming if you have some form of 3D or 2D cartesian system, where your quadrants are mapped on.
Although a little math heavy, it's up to you to how you would like to implement it in python (I would highly recommend creating multiple functions and optimising where you can (like the 4 rules to check intersection above)).
If you do not know any of the math concepts, google them to find out more, cause knowing how they work would help you a lot...

How to get the K most distant points, given their coordinates?

We have boring CSV with 10000 rows of ages (float), titles (enum/int), scores (float), ....
We have N columns each with int/float values in a table.
You can imagine this as points in ND space
We want to pick K points that would have maximised distance between each other.
So if we have 100 points in a tightly packed cluster and one point in the distance we would get something like this for three points:
or this
For 4 points it will become more interesting and pick some point in the middle.
So how to select K most distant rows (points) from N (with any complexity)? It looks like an ND point cloud "triangulation" with a given resolution yet not for 3d points.
I search for a reasonably fast approach (approximate - no precise solution needed) for K=200 and N=100000 and ND=6 (probably multigrid or ANN on KDTree based, SOM or triangulation based..).. Does anyone know one?
From past experience with a pretty similar problem, a simple solution of computing the mean Euclidean distance of all pairs within each group of K points and then taking the largest mean, works very well. As someone noted above, it's probably hard to avoid a loop on all combinations (not on all pairs). So a possible implementation of all this can be as follows:
import itertools
import numpy as np
from scipy.spatial.distance import pdist
Npoints = 3 # or 4 or 5...
# making up some data:
data = np.matrix([[3,2,4,3,4],[23,25,30,21,27],[6,7,8,7,9],[5,5,6,6,7],[0,1,2,0,2],[3,9,1,6,5],[0,0,12,2,7]])
# finding row indices of all combinations:
c = [list(x) for x in itertools.combinations(range(len(data)), Npoints )]
distances = []
for i in c:
distances.append(np.mean(pdist(data[i,:]))) # pdist: a method of computing all pairwise Euclidean distances in a condensed way.
ind = distances.index(max(distances)) # finding the index of the max mean distance
rows = c[ind] # these are the points in question
I propose an approximate solution. The idea is to start from a set of K points chosen in a way I'll explain below, and repeatedly loop through these points replacing the current one with the point, among the N-K+1 points not belonging to the set but including the current one, that maximizes the sum of the distances from the points of the set. This procedure leads to a set of K points where the replacement of any single point would cause the sum of the distances among the points of the set to decrease.
To start the process we take the K points that are closest to the mean of all points. This way we have good chances that on the first loop the set of K points will be spread out close to its optimum. Subsequent iterations will make adjustments to the set of K points towards a maximum of the sum of distances, which for the current values of N, K and ND appears to be reachable in just a few seconds. In order to prevent excessive looping in edge cases, we limit the number of loops nonetheless.
We stop iterating when an iteration does not improve the total distance among the K points. Of course, this is a local maximum. Other local maxima will be reached for different initial conditions, or by allowing more than one replacement at a time, but I don't think it would be worthwhile.
The data must be adjusted in order for unit displacements in each dimension to have the same significance, i.e., in order for Euclidean distances to be meaningful. E.g., if your dimensions are salary and number of children, unadjusted, the algorithm will probably yield results concentrated in the extreme salary regions, ignoring that person with 10 kids. To get a more realistic output you could divide salary and number of children by their standard deviation, or by some other estimate that makes differences in salary comparable to differences in number of children.
To be able to plot the output for a random Gaussian distribution, I have set ND = 2 in the code, but setting ND = 6, as per your request, is no problem (except you cannot plot it).
import matplotlib.pyplot as plt
import numpy as np
import scipy.spatial as spatial
N, K, ND = 100000, 200, 2
MAX_LOOPS = 20
SIGMA, SEED = 40, 1234
rng = np.random.default_rng(seed=SEED)
means, variances = [0] * ND, [SIGMA**2] * ND
data = rng.multivariate_normal(means, np.diag(variances), N)
def distances(ndarray_0, ndarray_1):
if (ndarray_0.ndim, ndarray_1.ndim) not in ((1, 2), (2, 1)):
raise ValueError("bad ndarray dimensions combination")
return np.linalg.norm(ndarray_0 - ndarray_1, axis=1)
# start with the K points closest to the mean
# (the copy() is only to avoid a view into an otherwise unused array)
indices = np.argsort(distances(data, data.mean(0)))[:K].copy()
# distsums is, for all N points, the sum of the distances from the K points
distsums = spatial.distance.cdist(data, data[indices]).sum(1)
# but the K points themselves should not be considered
# (the trick is that -np.inf ± a finite quantity always yields -np.inf)
distsums[indices] = -np.inf
prev_sum = 0.0
for loop in range(MAX_LOOPS):
for i in range(K):
# remove this point from the K points
old_index = indices[i]
# calculate its sum of distances from the K points
distsums[old_index] = distances(data[indices], data[old_index]).sum()
# update the sums of distances of all points from the K-1 points
distsums -= distances(data, data[old_index])
# choose the point with the greatest sum of distances from the K-1 points
new_index = np.argmax(distsums)
# add it to the K points replacing the old_index
indices[i] = new_index
# don't consider it any more in distsums
distsums[new_index] = -np.inf
# update the sums of distances of all points from the K points
distsums += distances(data, data[new_index])
# sum all mutual distances of the K points
curr_sum = spatial.distance.pdist(data[indices]).sum()
# break if the sum hasn't changed
if curr_sum == prev_sum:
break
prev_sum = curr_sum
if ND == 2:
X, Y = data.T
marker_size = 4
plt.scatter(X, Y, s=marker_size)
plt.scatter(X[indices], Y[indices], s=marker_size)
plt.grid(True)
plt.gca().set_aspect('equal', adjustable='box')
plt.show()
Output:
Splitting the data into 3 equidistant Gaussian distributions the output is this:
Assuming that if you read your csv file with N (10000) rows and D dimension (or features) into a N*D martix X. You can calculate the distance between each point and store it in a distance matrix as follows:
import numpy as np
X = np.asarray(X) ### convert to numpy array
distance_matrix = np.zeros((X.shape[0],X.shape[0]))
for i in range(X.shape[0]):
for j in range(i+1,X.shape[0]):
## We compute triangle matrix and copy the rest. Distance from point A to point B and distance from point B to point A are the same.
distance_matrix[i][j]= np.linalg.norm(X[i]-X[j]) ## Here I am calculating Eucledian distance. Other distance measures can also be used.
#distance_matrix = distance_matrix + distance_matrix.T - np.diag(np.diag(distance_matrix)) ## This syntax can be used to get the lower triangle of distance matrix, which is not really required in your case.
K = 5 ## Number of points that you want to pick
indexes = np.unravel_index(np.argsort(distance_matrix.ravel())[-1*K:], distance_matrix.shape)
print(indexes)
Bottom Line Up Front: Dealing with multiple equally distant points and the Curse of Dimensionality are going to be larger problems than just finding the points. Spoiler alert: There's a surprise ending.
I think this an interesting question but I'm bewildered by some of the answers. I think this is, in part, due to the sketches provided. You've no doubt noticed the answers look similar -- 2d, with clusters -- even though you indicated a wider scope was needed. Because others will eventually see this, I'm going to step through my thinking a bit slowly so bear with me for the early part.
It makes sense to start with a simplified example to see if we can generalize a solution with data that's easy to grasp and a linear 2D model is easiest of the easy.
We don't need to calculate all the distances though. We just need the ones at the extremes. So we can then take the top and bottom few values:
right = lin_2_D.nlargest(8, ['x'])
left = lin_2_D.nsmallest(8, ['x'])
graph = sns.scatterplot(x="x", y="y", data=lin_2_D, color = 'gray', marker = '+', alpha = .4)
sns.scatterplot(x = right['x'], y = right['y'], color = 'red')
sns.scatterplot(x = left['x'], y = left['y'], color = 'green')
fig = graph.figure
fig.set_size_inches(8,3)
What we have so far: Of 100 points, we've eliminated the need to calculate the distance between 84 of them. Of what's left we can further drop this by ordering the results on one side and checking the distance against the others.
You can imagine a case where you have a couple of data points way off the trend line that could be captured by taking the greatest or least y values, and all that starts to look like Walter Tross's top diagram. Add in a couple of extra clusters and you get what looks his bottom diagram and it appears that we're sort of making the same point.
The problem with stopping here is the requirement you mentioned is that you need a solution that works for any number of dimensions.
The unfortunate part is that we run into four challenges:
Challenge 1: As you increase the dimensions you can run into a large number of cases where you have multiple solutions when seeking midpoints. So you're looking for k furthest points but have a large number of equally valid possible solutions and no way prioritizing them. Here are two super easy examples illustrate this:
A) Here we have just four points and in only two dimensions. You really can't get any easier than this, right? The distance from red to green is trivial. But try to find the next furthest point and you'll see both of the black points are equidistant from both the red and green points. Imagine you wanted the furthest six points using the first graphs, you might have 20 or more points that are all equidistant.
edit: I just noticed the red and green dots are at the edges of their circles rather than at the center, I'll update later but the point is the same.
B) This is super easy to imagine: Think of a D&D 4 sided die. Four points of data in a three-dimensional space, all equidistant so it's known as a triangle-based pyramid. If you're looking for the closest two points, which two? You have 4 choose 2 (aka, 6) combinations possible. Getting rid of valid solutions can be a bit of a problem because invariably you face questions such as "why did we get rid of these and not this one?"
Challenge 2: The Curse of Dimensionality. Nuff Said.
Challenge 3 Revenge of The Curse of Dimensionality Because you're looking for the most distant points, you have to x,y,z ... n coordinates for each point or you have to impute them. Now, your data set is much larger and slower.
Challenge 4 Because you're looking for the most distant points, dimension reduction techniques such as ridge and lasso are not going to be useful.
So, what to do about this?
Nothing.
Wait. What?!?
Not truly, exactly, and literally nothing. But nothing crazy. Instead, rely on a simple heuristic that is understandable and computationally easy. Paul C. Kainen puts it well:
Intuitively, when a situation is sufficiently complex or uncertain,
only the simplest methods are valid. Surprisingly, however,
common-sense heuristics based on these robustly applicable techniques
can yield results which are almost surely optimal.
In this case, you have not the Curse of Dimensionality but rather the Blessing of Dimensionality. It's true you have a lot of points and they'll scale linearly as you seek other equidistant points (k) but the total dimensional volume of space will increase to power of the dimensions. The k number of furthest points you're is insignificant to the total number of points. Hell, even k^2 becomes insignificant as the number of dimensions increase.
Now, if you had a low dimensionality, I would go with them as a solution (except the ones that are use nested for loops ... in NumPy or Pandas).
If I was in your position, I'd be thinking how I've got code in these other answers that I could use as a basis and maybe wonder why should I should trust this other than it lays out a framework on how to think through the topic. Certainly, there should be some math and maybe somebody important saying the same thing.
Let me reference to chapter 18 of Computer Intensive Methods in Control and Signal Processing and an expanded argument by analogy with some heavy(-ish) math. You can see from the above (the graph with the colored dots at the edges) that the center is removed, particularly if you followed the idea of removing the extreme y values. It's a though you put a balloon in a box. You could do this a sphere in a cube too. Raise that into multiple dimensions and you have a hypersphere in a hypercube. You can read more about that relationship here.
Finally, let's get to a heuristic:
Select the points that have the most max or min values per dimension. When/if you run out of them pick ones that are close to those values if there isn't one at the min/max. Essentially, you're choosing the corners of a box For a 2D graph you have four points, for a 3D you have the 8 corners of the box (2^3).
More accurately this would be a 4d or 5d (depending on how you might assign the marker shape and color) projected down to 3d. But you can easily see how this data cloud gives you the full range of dimensions.
Here is a quick check on learning; for purposes of ease, ignore the color/shape aspect: It's easy to graphically intuit that you have no problem with up to k points short of deciding what might be slightly closer. And you can see how you might need to randomize your selection if you have a k < 2D. And if you added another point you can see it (k +1) would be in a centroid. So here is the check: If you had more points, where would they be? I guess I have to put this at the bottom -- limitation of markdown.
So for a 6D data cloud, the values of k less than 64 (really 65 as we'll see in just a moment) points are pretty easy. But...
If you don't have a data cloud but instead have data that has a linear relationship, you'll 2^(D-1) points. So, for that linear 2D space, you have a line, for linear 3D space, you'd have a plane. Then a rhomboid, etc. This is true even if your shape is curved. Rather than do this graph myself, I'm using the one from an excellent post on by Inversion Labs on Best-fit Surfaces for 3D Data
If the number of points, k, is less than 2^D you need a process to decide what you don't use. Linear discriminant analysis should be on your shortlist. That said, you can probably satisfice the solution by randomly picking one.
For a single additional point (k = 1 + 2^D), you're looking for one that is as close to the center of the bounding space.
When k > 2^D, the possible solutions will scale not geometrically but factorially. That may not seem intuitive so let's go back to the two circles. For 2D you have just two points that could be a candidate for being equidistant. But if that were 3D space and rotate the points about the line, any point in what is now a ring would suffice as a solution for k. For a 3D example, they would be a sphere. Hyperspheres (n-spheres) from thereon. Again, 2^D scaling.
One last thing: You should seriously look at xarray if you're not already familiar with it.
Hope all this helps and I also hope you'll read through the links. It'll be worth the time.
*It would be the same shape, centrally located, with the vertices at the 1/3 mark. So like having 27 six-sided dice shaped like a giant cube. Each vertice (or point nearest it) would fix the solution. Your original k+1 would have to be relocated too. So you would select 2 of the 8 vertices. Final question: would it be worth calculating the distances of those points against each other (remember the diagonal is slightly longer than the edge) and then comparing them to the original 2^D points? Bluntly, no. Satifice the solution.
If you're interested in getting the most distant points you can take advantage of all of the methods that were developed for nearest neighbors, you just have to give a different "metric".
For example, using scikit-learn's nearest neighbors and distance metrics tools you can do something like this
import numpy as np
from sklearn.neighbors import BallTree
from sklearn.neighbors.dist_metrics import PyFuncDistance
from sklearn.datasets import make_blobs
from matplotlib import pyplot as plt
def inverted_euclidean(x1, x2):
# You can speed this up using cython like scikit-learn does or numba
dist = np.sum((x1 - x2) ** 2)
# We invert the euclidean distance and set nearby points to the biggest possible
# positive float that isn't inf
inverted_dist = np.where(dist == 0, np.nextafter(np.inf, 0), 1 / dist)
return inverted_dist
# Make up some fake data
n_samples = 100000
n_features = 200
X, _ = make_blobs(n_samples=n_samples, centers=3, n_features=n_features, random_state=0)
# We exploit the BallTree algorithm to get the most distant points
ball_tree = BallTree(X, leaf_size=50, metric=PyFuncDistance(inverted_euclidean))
# Some made up query, you can also provide a stack of points to query against
test_point = np.zeros((1, n_features))
distance, distant_points_inds = ball_tree.query(X=test_point, k=10, return_distance=True)
distant_points = X[distant_points_inds[0]]
# We can try to visualize the query results
plt.plot(X[:, 0], X[:, 1], ".b", alpha=0.1)
plt.plot(test_point[:, 0], test_point[:, 1], "*r", markersize=9)
plt.plot(distant_points[:, 0], distant_points[:, 1], "sg", markersize=5, alpha=0.8)
plt.show()
Which will plot something like:
There are many points that you can improve on:
I implemented the inverted_euclidean distance function with numpy, but you can try to do what the folks of scikit-learn do with their distance functions and implement them in cython. You could also try to jit compile them with numba.
Maybe the euclidean distance isn't the metric you would like to use to find the furthest points, so you're free to implement your own or simply roll with what scikit-learn provides.
The nice thing about using the Ball Tree algorithm (or the KdTree algorithm) is that for each queried point you have to do log(N) comparisons to find the furthest point in the training set. Building the Ball Tree itself, I think also requires log(N) comparison, so in the end if you want to find the k furthest points for every point in the ball tree training set (X), it will have almost O(D N log(N)) complexity (where D is the number of features), which will increase up to O(D N^2) with the increasing k.

How can I correct the normal vectors of 3D model? [duplicate]

I'm a newbie to computer graphics so I apologize if some of my language is inexact or the question misses something basic.
Is it possible to calculate face normals correctly, given a list of vertices, and a list of faces like this:
v1: x_1, y_1, z_1
v2: x_2, y_2, z_2
...
v_n: x_n, y_n, z_n
f1: v1,v2,v3
f2: v4,v2,v5
...
f_m: v_j, v_k, v_l
Each x_i, y_i , z_i specifies the vertices position in 3d space (but isn't neccesarily a vector)
Each f_i contains the indices of the three vertices specifying it.
I understand that you can use the cross product of two sides of a face to get a normal, but the direction of that normal depends on the order and choice of sides (from what I understand).
Given this is the only data I have is it possible to correctly determine the direction of the normals? or is it possible to determine them consistently atleast? (all normals may be pointing in the wrong direction?)
In general there is no way to assign normal "consistently" all over a set of 3d faces... consider as an example the famous Möbius strip...
You will notice that if you start walking on it after one loop you get to the same point but on the opposite side. In other words this strip doesn't have two faces, but only one. If you build such a shape with a strip of triangles of course there's no way to assign normals in a consistent way and you'll necessarily end up having two adjacent triangles with normals pointing in opposite directions.
That said, if your collection of triangles is indeed orientable (i.e. there actually exist a consistent normal assignment) a solution is to start from one triangle and then propagate to neighbors like in a flood-fill algorithm. For example in Python it would look something like:
active = [triangles[0]]
oriented = set([triangles[0]])
while active:
next_active = []
for tri in active:
for other in neighbors(tri):
if other not in oriented:
if not agree(tri, other):
flip(other)
oriented.add(other)
next_active.append(other)
active = next_active
In CG its done by polygon winding rule. That means all the faces are defined so the points are in CW (or CCW) order when looked on the face directly. Then using cross product will lead to consistent normals.
However many meshes out there does not comply the winding rule (some faces are CW others CCW not all the same) and for those its a problem. There are two approaches I know of:
for simple shapes (not too much concave)
the sign of dot product of your face_normal and face_center-cube_center will tell you if the normal points inside or outside of the object.
if ( dot( face_normal , face_center-cube_center ) >= 0.0 ) normal_points_out
You can even use any point of face instead of the face center too. Anyway for more complex concave shapes this will not work correctly.
test if point above face is inside or not
simply displace center of face by some small distance (not too big) in normal direction and then test if the point is inside polygonal mesh or not:
if ( !inside( face_center+0.001*face_normal ) ) normal_points_out
to check if point is inside or not you can use hit test.
However if the normal is used just for lighting computations then its usage is usually inside a dot product. So we can use its abs value instead and that will solve all lighting problems regardless of the normal side. For example:
output_color = face_color * abs(dot(face_normal,light_direction))
some gfx apis have implemented this already (look for double sided materials or normals, turning them on usually use the abs value ...) For example in OpenGL:
glLightModeli(GL_LIGHT_MODEL_TWO_SIDE, GL_TRUE);

minimize euclidean distance from sets of points in n-dimensions

Let's look at m points in n-d space- (A solution for 4 points in 3-d space is here: minimize distance from sets of points)
a= (x1, y1, z1, ..)
b= (x2, y2 ,z2, ..)
c= (x3, y3, z3, ..)
.
.
p= (x , y , z, ..)
Find point q = c1* a + c2* b + c3* c + ..
where c1 + c2 + c3 + .. = 1
and c1, c2, c3, .. >= 0
s.t.
euclidean distance pq is minimized.
What algorithms can be used ? Idea or pseudocode is enough.
(Optimizing performance is a big issue here. Monte Carlo method with all vertices and changing coefficients would also give a solution.)
We can assume p = 0 by subtracting p from all the other points. Then the question is one of minimizing the norm over a convex hull of a finite set of points, i.e., a polytope.
There are a few papers on this problem. It looks like "A recursive algorithm for finding the minimum norm point in a polytope and a pair of closest points in two polytopes" by Kazuyuki Sekitani and Yoshitsugu Yamamoto is a good one, with a short survey of prior solutions to the problem. It is behind a paywall but if you have access to a university library you may be able to download a copy.
The algorithm they give is fairly simple, once you get past the notation. P is the finite set of points. C(P) is its convex hull. Nr(C(P)) is the unique point of minimum norm, which is what you want to find.
Step 0: Choose a point x_0 from the convex hull C(P) of your finite set of points P. They recommend choosing x_0 to be the point in P with minimum norm. Let k=1.
Now loop:
Step 1: Let a_k = min {x^t_{k-1} p | p is in P}. Here x^t_{k-1} is the transpose of x_{k-1} (so the function being minimized is just a dot product as p ranges over your finite set P). If |x_{k-1}|^2 <= a_k, then the answer is x_{k-1}, stop.
Step 2: P_k = {p | p in P and x^t_{k-1} = a_k}. P_k is the subset of P that minimizes the expression in Step 1. Call the algorithm recursively on this set P_k, and let the result be y_k = Nr(C(P_k)).
Step 3: b_k = min{y^t_k p | p in P\P_k}, the minimum of the dot product of y_k with points in the complement set P\P_k. If |y_k|^2 <= b_k then y_k is the answer, stop.
Step 4: s_k = max{s| [(1-s)x_{k-1} + sy_k]^t y_k <= [(1-s)x_{k-1} + sy_k]^t p for every p in P\P_k}. Let x_k = (1-s_k) x_{k-1} + s_k y_k, let k=k+1, and go back to Step 1.
There is an explicit formula for s_k in Step 4:
s_k = min{ [x^t_{k-1} (p-y_k)]/[(y_k-x_{k-1})^t (y_k-p)] | p in P\P_k and (y_k - x_{k-1})^t (y_k-p) > 0 }
There is a proof in the paper that s_k has the necessary properties, that the algorithm terminates after a finite number of operations, and that the result is indeed optimal.
Note that you should add some tolerance into your comparisons, otherwise rounding errors may cause the algorithm to fail. There is a lot of discussion about numerical stability, see the paper for details.
They do not give a complete analysis of the computational complexity of the algorithm, but they do prove it is at most O(m^2) in the two-dimensional case (m is the number of points in P), and they have done numerical experiments which give the impression that it is sublinear in time as a function of m, with dimension fixed. I'm skeptical of that claim. In the absence of a detailed analysis, I suggest you try some experiments with typical data to see how well the algorithm performs for you.
Stated a simpler way, you have a set of points {a}i, and you are considering all points which are some weighted average thereof. This set of points is exactly the convex hull of those points; it's a polytope (polygon, polyhedron, etc.) that just happens to be convex, where the corners are a subset of the {a}i points.
You are just asking which point on a polytope(~hedron) is closest to a point. (your query point p)
The closest point must be on the exterior of the polytope. One algorithm would be to brute-force searching all N-1 dimensional surfaces. Do this in the usual way you would find the closest point on a line or surface or N-dimensional surface to a query point.
(If the points are not all linearly independent, you will have multiple ways (multiple weight vectors) which can give you the same weighted-average point q. You can worry about reconstructing the answer q from the basis vectors after you find it geometrically.)

Calculating the intersection area between two rectangles with axes not aligned

I want to calculate the intersection over union IoU between two rectangles with axes not aligned, but with an angle of the axes smaller than 30 degrees. An approximate value is also seeked.
One possible solution is to check if the angle between the two rectangles is less than 30 degree and than rotate them parallel to aligne the axis. From here it is easy to calculate the IoU.
Another possibility is to use monte carlo methods for the intersection ( generate a point, find if the point is under some line of one rectangle and above some line of the other), but this seems expensive because I need to use this calculation a large number of times.
I was hopping that there is something better out there; maybe a geometry library, or maybe an algorithm from the computer vision folks.
I am trying to learn grasping positions using deep neural networks. My algorithem should predict a bounding box (rectangle) for an object in an rgb image. For any image I have also the ground truth (another rectangle) bounding box. From this two rectangles I need the IoU.
Any idea?
Since you're working in Python, I think the Shapely package would serve your needs.
There is quite effective algorithm for calculation of intersection between two convex polygons, described in O'Rourke book "Computational Geometry in C".
C code is available at the book page (convconv).
Algorithm traverse edges of the first polygon, checking orientations of the second polygon vertices in order to detect intersections. When two consequent vertices lie on the different sides of the edge, intersection occurs (there is a lot of trick cases). Algorithm outline is here
You can consider a number of numerical approaches, practically "rendering" the rectangles into some "canvas"/canvases, and traverse the pixels for making your statistics. The size of the canvas should be the size of the bounding box for the entire scene, practically you can find that via picking the minimum and maximum coordinates occurring for each axis.
1) "most CG" approach: really get a rendering library, render one rectangle with red, other rectangle with transparent blue. Then visit each pixel and if it has a non-0 red component, it belongs to the first rectangle, if it has a non-0 blue component, it belongs to the second rectangle. And if it has both, it belongs to the intersection too. This approach is cheap for coding, but requires both writing and reading the canvas even in the rendering phase, which is slower than just writing. This might be even done on GPU too, though I am not sure if setup costs and getting back the result do not weight out the benefit for such a simple scene.
2) another CG-approach would be rendering into 2 arrays, preferably some 1-byte-per-pixel variant, for the sake of speed (you may have to go back in time a bit in order to find such dedicated rendering libraries). This way the renderer only writes, into one array per rectangle, and you read from two when creating the statistics
3) as writing and reading pixels take time, you can do some shortcut, but it needs more coding: convex shapes can be rendered via collecting the minimum and maximum coordinates per scanline, and just filling between the two. If you do it yourself, you can spare the filling part and also the read-and-check-every-pixel step at the end. Build such min-max list for both rectangles, and then you "just" have to check their relation/order for each scanline, to recognize overlaps
And then there is the mathematical way: this is not really useful, see EDIT below while it is unlikely that you would find some sane algorithm for calculating intersection area, specifically for the case of rectangles, if you find such algorithm for triangles, which is more probable, that would be enough. Both rectangles can be split into two triangles, 1A+1B and 2A+2B respectively, and then you just have to run such algorithm 4 times: 1A-2A, 1A-2B, 1B-2A, 1B-2B, sum the results and that is the area of your intersection.
EDIT: for the maths approach (though this also comes from graphics), I think https://en.wikipedia.org/wiki/Sutherland%E2%80%93Hodgman_algorithm can be applied here (as both rectangles are convex polygons, A-B and B-A should produce the same result) for finding the intersection polygon, and then the remaining task is to calculate the area of that polygon (here and now I think it is going to be convex, and then it is really easy).
I ended up using Sutherland-Hodgman algorithm implemented as this functions:
def clip(subjectPolygon, clipPolygon):
def inside(p):
return(cp2[0]-cp1[0])*(p[1]-cp1[1]) > (cp2[1]-cp1[1])*(p[0]-cp1[0])
def computeIntersection():
dc = [ cp1[0] - cp2[0], cp1[1] - cp2[1] ]
dp = [ s[0] - e[0], s[1] - e[1] ]
n1 = cp1[0] * cp2[1] - cp1[1] * cp2[0]
n2 = s[0] * e[1] - s[1] * e[0]
n3 = 1.0 / (dc[0] * dp[1] - dc[1] * dp[0])
return [(n1*dp[0] - n2*dc[0]) * n3, (n1*dp[1] - n2*dc[1]) * n3]
outputList = subjectPolygon
cp1 = clipPolygon[-1]
for clipVertex in clipPolygon:
cp2 = clipVertex
inputList = outputList
outputList = []
s = inputList[-1]
for subjectVertex in inputList:
e = subjectVertex
if inside(e):
if not inside(s):
outputList.append(computeIntersection())
outputList.append(e)
elif inside(s):
outputList.append(computeIntersection())
s = e
cp1 = cp2
return(outputList)
def PolygonArea(corners):
n = len(corners) # of corners
area = 0.0
for i in range(n):
j = (i + 1) % n
area += corners[i][0] * corners[j][1]
area -= corners[j][0] * corners[i][1]
area = abs(area) / 2.0
return area
intersection = clip(rec1, rec2)
intersection_area = PolygonArea(intersection)
iou = intersection_area/(PolygonArea(rec1)+PolygonArea(rec2)-intersection_area)
Another slower method (don't know what algorithm) could be:
from shapely.geometry import Polygon
p1 = Polygon(rec1)
p2 = Polygon(rec2)
inter_sec_area = p1.intersection(rec2).area
iou = inter_sec_area/(p1.area + p2.area - inter_sec_area)
It is worth mentioning that in just one case with bigger polygons (not my case) the shapely module had an area twice greater than the first method. I didn't test both methods intensively.
This might help
What about using Pythagorean theorem ? Since you have two rectangles, when they intersect, you will have one or more triangles, each with one angle of 90°.
Since you also know the angle between the two rectangles (20° in my example), and the coordinates of each rectangle, you can use the the appropriate function (cos/sin/tan) to know the length of all the edges of the triangles.
I hope this can help

Categories