Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 3 years ago.
Improve this question
I have many images consisting of circles(varying from 1 to 4) in each image. I am trying to get a clear circle images by filling the missed pixels along the circle path.
I have tried Hough-transform, but its parameters are image specific: for each image I have to change the parameters. With this problem I am unable keep them in a single for loop.
Please provide some method to do it. Thanks
imfindcircles Does not work
The most "natural" way to approach this problem is to use Matlab's imfindcircles. However, that function assume the circles in the image are "full", yet in your examples you only have the (incomplete) boundaries of the circles, thus imfindcircles cannot be directly applied to your data.
Alternative Approach
You can use ransac to fit circles to your data. Fit one circle at a time to as many points as you can, terminate when there are too few points left that fit no circle at all.
To use RanSac you basically need to implement two methods:
Model fitting method, fitFcn, Given a small sample of your points - fit a circle to them.
Distance to model method, distFcn, Given a circle ("model") find the distance of each point to that circle.
Once you have these two methods, RanSac operates roughly like:
- randomly sample very few points
- use fitFcn to fit a circle to these sampled points
- use distFcn to compute the distance of all points to estimated circle
- if enough points are close to the circle, accept this circle and remove all point that "belongs" to that circle
- terminate if no circle was found or not enough "unexplained" points
This can be easily implemented in Matlab.
First consider fitFcn: we need a function that compute (cx, cy, r) - the three parameters of a 2D circle (center and radii). Given a point (x, y) it fits a circle iff
(x - cx)^2 + (y - cy)^2 = r^2
We can write this equation as a linear relation between known points (x, y) and unknown circle (cx, cy, r) in the following manner
[-2*x, -2*y, 1] [ cx ;
cy ; = [-x^2-y^2]
cx^2 + cy^2 - r^2 ]
Using a least squares estimation (in a similar manner as in this answer), we can recover the circle parameters given enough points (at least 3) on the circle
This is how the code actually looks like
function crc = fit_circle(xy) % xy is n-by-2 matrix of point coordinates
% fit in least squares sens
x = xy(:, 1);
y = xy(:, 2);
X = [-2*x, -2*y, ones(size(x))];
Y = -x.^2 - y.^2;
crc = (X\Y).'; % least squares solution
r2 = -crc(3) +crc(1).^2 + crc(2).^2;
if r2 <= 0
crc(3) = 0;
else
crc(3) = sqrt(r2);
end
% output crc is a 3 vector (cx, cy, r)
Now that we can fit a circle to points, we need to compute the distance using distFcn that is quite simple
function dst = distFcn(crc, xy)
% how good a fit circle for points
x = xy(:, 1) - crc(1);
y = xy(:, 2) - crc(2);
dst = abs(sqrt(x.^2 + y.^2) - crc(3));
Putting it all together with matlab's ransac:
function circles = find_circles(bw)
% parameters
sample_size = 4;
max_distance = 10;
min_num_points_in_circle = 50;
[y, x] = find(bw > max(bw(:))/2); % all edges in the image
circles = {};
counter = 0;
while numel(x) > 10 * sample_size && counter < 10
try
[circle, inlierIdx] = ransac([x, y], #fit_circle, #distFcn, ...
sample_size, max_distance);
catch
break
end
% refit using only inliers
circle = fit_circle([x(inlierIdx) y(inlierIdx)]);
dst = distFcn(circle, [x y]);
founfit = dst < max_distance;
if sum(founfit) > min_num_points_in_circle
% this model fits enough points
circles{end+1} = circle;
x(founfit) = [];
y(founfit) = [];
else
counter = counter + 1;
end
end
circles = vertcat(circles{:});
And the output of this function on your data is (using viscircles to plot the circles):
Related
Background :
I am implementing an application where I need to do affine transformations on a selected line. Line is represented using 2 points. All the code is done in mouse events.
void mousePressEvent(){
lastPoint = event.point();
}
void mouseMoveEvent(){
currentPoint = event.point();
//Do transformations here.
translate_factor = currentPoint - lastPoint
scale_factor = currentPoint / lastPoint
rotation_angle = f(current_point,last_point) //FIND f
//Apply transformation
lastPoint = currentPoint
}
I have two points, p1 and p2. The X axis increases from left to right. Y axis increases top to bottom. I have a point d around which I want to rotate a set of points p_set .
How do I find the angle of rotation from p1 to p2 about d .
What I am doing is translate p1 , p2 by -d and calculate theta by using atan2f as shown in the following code.
Note that y is inverted : goes from top to bottom.
Here is the code
const auto about = stroke->getMidPoint();
Eigen::Affine2f t1(Eigen::Translation2f(0-about.x(),0-about.y()));
Eigen::Vector3f p1 = t1.matrix()*Eigen::Vector3f(lastPoint.x(),lastPoint.y(),1.0);
Eigen::Vector3f p2 = t1.matrix()*Eigen::Vector3f(currentPoint.x(),currentPoint.y(),1.0);
float theta = atan2f(p2[1]-p1[1],p2[0]-p1[0]);
When I transform using this, I get very weird rotations.
What I want is to find angle as a function of current and last points
It seems to me that you're finding the slope of the vector p2-p1.
If I understood correctly, you want the difference of the slopes of the two vectors p2-d and p1-d, so the rotation angle is something like atan((p2.y-d.y)/(p2.x-d.x)) - atan((p1.y-d.y)/(p1.x-d.x))
I've got a shape consisting of four points, A, B, C and D, of which the only their position is known. The goal is to transform these points to have specific angles and offsets relative to each other.
For example: A(-1,-1) B(2,-1) C(1,1) D(-2,1), which should be transformed to a perfect square (all angles 90) with offsets between AB, BC, CD and AD all being 2. The result should be a square slightly rotated counter-clockwise.
What would be the most efficient way to do this?
I'm using this for a simple block simulation program.
As Mark alluded, we can use constrained optimization to find the side 2 square that minimizes the square of the distance to the corners of the original.
We need to minimize f = (a-A)^2 + (b-B)^2 + (c-C)^2 + (d-D)^2 (where the square is actually a dot product of the vector argument with itself) subject to some constraints.
Following the method of Lagrange multipliers, I chose the following distance constraints:
g1 = (a-b)^2 - 4
g2 = (c-b)^2 - 4
g3 = (d-c)^2 - 4
and the following angle constraints:
g4 = (b-a).(c-b)
g5 = (c-b).(d-c)
A quick napkin sketch should convince you that these constraints are sufficient.
We then want to minimize f subject to the g's all being zero.
The Lagrange function is:
L = f + Sum(i = 1 to 5, li gi)
where the lis are the Lagrange multipliers.
The gradient is non-linear, so we have to take a hessian and use multivariate Newton's method to iterate to a solution.
Here's the solution I got (red) for the data given (black):
This took 5 iterations, after which the L2 norm of the step was 6.5106e-9.
While Codie CodeMonkey's solution is a perfectly valid one (and a great use case for the Lagrangian Multipliers at that), I believe that it's worth mentioning that if the side length is not given this particular problem actually has a closed form solution.
We would like to minimise the distance between the corners of our fitted square and the ones of the given quadrilateral. This is equivalent to minimising the cost function:
f(x1,...,y4) = (x1-ax)^2+(y1-ay)^2 + (x2-bx)^2+(y2-by)^2 +
(x3-cx)^2+(y3-cy)^2 + (x4-dx)^2+(y4-dy)^2
Where Pi = (xi,yi) are the corners of the fitted square and A = (ax,ay) through D = (dx,dy) represent the given corners of the quadrilateral in clockwise order. Since we are fitting a square we have certain contraints regarding the positions of the four corners. Actually, if two opposite corners are given, they are enough to describe a unique square (save for the mirror image on the diagonal).
Parametrization of the points
This means that two opposite corners are enough to represent our target square. We can parametrise the two remaining corners using the components of the first two. In the above example we express P2 and P4 in terms of P1 = (x1,y1) and P3 = (x3,y3). If you need a visualisation of the geometrical intuition behind the parametrisation of a square you can play with the interactive version.
P2 = (x2,y2) = ( (x1+x3-y3+y1)/2 , (y1+y3-x1+x3)/2 )
P4 = (x4,y4) = ( (x1+x3+y3-y1)/2 , (y1+y3+x1-x3)/2 )
Substituting for x2,x4,y2,y4 means that f(x1,...,y4) can be rewritten to:
f(x1,x3,y1,y3) = (x1-ax)^2+(y1-ay)^2 + ((x1+x3-y3+y1)/2-bx)^2+((y1+y3-x1+x3)/2-by)^2 +
(x3-cx)^2+(y3-cy)^2 + ((x1+x3+y3-y1)/2-dx)^2+((y1+y3+x1-x3)/2-dy)^2
a function which only depends on x1,x3,y1,y3. To find the minimum of the resulting function we then set the partial derivatives of f(x1,x3,y1,y3) equal to zero. They are the following:
df/dx1 = 4x1-dy-dx+by-bx-2ax = 0 --> x1 = ( dy+dx-by+bx+2ax)/4
df/dx3 = 4x3+dy-dx-by-bx-2cx = 0 --> x3 = (-dy+dx+by+bx+2cx)/4
df/dy1 = 4y1-dy+dx-by-bx-2ay = 0 --> y1 = ( dy-dx+by+bx+2ay)/4
df/dy3 = 4y3-dy-dx-2cy-by+bx = 0 --> y3 = ( dy+dx+by-bx+2cy)/4
You may see where this is going, as simple rearrangment of the terms leads to the final solution.
Final solution
is there some good and better way to find centroid of contour in opencv, without using built in functions?
While Sonaten's answer is perfectly correct, there is a simple way to do it: Use the dedicated opencv function for that: moments()
http://opencv.itseez.com/modules/imgproc/doc/structural_analysis_and_shape_descriptors.html?highlight=moments#moments
It does not only returns the centroid, but some more statistics about your shape. And you can send it a contour or a raster shape (binary image), whatever best fits your need.
EDIT
example (modified) from "Learning OpenCV", by gary bradsky
CvMoments moments;
double M00, M01, M10;
cvMoments(contour,&moments);
M00 = cvGetSpatialMoment(&moments,0,0);
M10 = cvGetSpatialMoment(&moments,1,0);
M01 = cvGetSpatialMoment(&moments,0,1);
centers[i].x = (int)(M10/M00);
centers[i].y = (int)(M01/M00);
What you get in your current piece of code is of course the centroid of your bounding box.
"If you have a bunch of points(2d vectors), you should be able to get the centroid by averaging those points: create a point to add all the other points' positions into and then divide the components of that point with accumulated positions by the total number of points." - George Profenza mentions
This is indeed the right approach for the exact centroid of any given object in two-dimentionalspace.
On wikipedia we have some general forms for finding the centroid of an object.
http://en.wikipedia.org/wiki/Centroid
Personally, I would ask myself what I needed from this program. Do I want a thorough but performance heavy operation, or do I want to make some approximations? I might even be able to find an OpenCV function that deals with this correct and efficiently.
Don't have a working example, so I'm writing this in pseudocode on a simple 5 pixel example on a thorough method.
x_centroid = (pixel1_x + pixel2_x + pixel3_x + pixel4_x +pixel5_x)/5
y_centroid = (pixel1_y + pixel2_y + pixel3_y + pixel4_y +pixel5_y)/5
centroidPoint(x_centroid, y_centroid)
Looped for x pixels
Loop j times *sample (for (int i=0, i < j, i++))*
{
x_centroid = pixel[j]_x + x_centroid
y_centroid = pixel[j]_x + x_centroid
}
x_centroid = x_centroid/j
y_centroid = y_centroid/j
centroidPoint(x_centroid, y_centroid)
Essentially, you have the vector contours of the type
vector<vector<point>>
in OpenCV 2.3. I believe you have something similar in earlier versions, and you should be able to go through each blob on your picture with the first index of this "double vector", and go through each pixel in the inner vector.
Here is a link to documentation on the contour function
http://opencv.itseez.com/modules/imgproc/doc/structural_analysis_and_shape_descriptors.html?highlight=contours#cv.DrawContours
note: you've tagged your question as c++ visual. I'd suggest that you use the c++ syntax in OpenCV 2.3 instead of c. The first and good reason to use 2.3 is that it is more class based, which in this case means that the class Mat (instead of IplImage) does leak memory. One does not have to write destroy commands all the live long day :)
I hope this shed some light on your problem. Enjoy.
I've used Joseph O'Rourke excellent polygon centroid algorithm to great success.
See http://maven.smith.edu/~orourke/Code/centroid.c
Essentially:
For each point in the contour, find the triangle area from the current index polygon xy to the next 2 polygon xy points e.g.: Math.Abs(((X1 - X0) * (Y2 - Y0) - (X2 - X0) * (Y1 - Y0)) / 2)
Add this triangle area to a list TriAreas
Sum the triangle area, and store in SumT
Find the centroid CTx and CTy from this current triangle: CTx = (X0 + X1 + X2) / 3 and CTy = (Y0 + Y1 + Y2) / 3;
Store these 2 centroid values in 2 other lists CTxs CTys.
Finally after performing this with all points in the contour, find the contours centroid x and y using the 2 triangle x and y lists in 5 which is a weighted sum of signed triangle areas, weighted by the centroid of each triangle:
for (Int32 Index = 0; Index < CTxs.Count; Index++)
{
CentroidPointRet.X += CTxs[Index] * (TriAreas[Index] / SumT);
}
// now find centroid Y value
for (Int32 Index = 0; Index < CTys.Count; Index++)
{
CentroidPointRet.Y += CTys[Index] * (TriAreas[Index] / SumT);
}
Imagine a circle. Imagine a pie. Imagine trying to return a bool that determines whether the provided parameters of X, Y are contained within one of those pie pieces.
What I know about the arc:
I have the CenterX, CenterY, Radius, StartingAngle, EndingAngle, StartingPoint (point on circumference), EndingPoint (point on circumference).
Given a coordinate of X,Y, I'd like to determine if this coordinate is contained anywhere within the pie slide.
Check:
The angle from the centerX,centerY through X,Y should be between start&endangle.
The distance from centerX,centerY to X,Y should be less then the Radius
And you'll have your answer.
I know this question is old but none of the answers consider the placement of the arc on the circle.
This algorithm considers that all angles are between 0 and 360, and the arcs are drawn in positive mathematical direction (counter-clockwise)
First you can transform to polar coordinates: radius (R) and angle (A). Note: use Atan2 function if available. wiki
R = sqrt ((X - CenterX)^2 + (Y - CenterY)^2)
A = atan2 (Y - CenterY, X - CenterX)
Now if R < Radius the point is inside the circle.
To check if the angle is between StartingAngle (S) and EndingAngle (E) you need to consider two possibilities:
1) if S < E then if S < A < E the point lies inside the slice
2) if S > E then there are 2 possible scenarios
if A > S
then the point lies inside the slice
if A < E
then the point lies inside the slice
In all other cases the point lies outside the slice.
Convert X,Y to polar coordinates using this:
Angle = arctan(y/x);
Radius = sqrt(x * x + y * y);
Then Angle must be between StartingAngle and EndingAngle, and Radius between 0 and your Radius.
You have to convert atan2() to into 0-360 before making comparisons with starting and ending angles.
(A > 0 ? A : (2PI + A)) * 360 / (2PI)
I have read similar topics in order to find solution, but with no success.
What I'm trying to do is make the tool same as can be found in CorelDraw, named "Pen Tool". I did it by connecting Bezier cubic curves, but still missing one feature, which is dragging curve (not control point) in order to edit its shape.
I can successfully determine the "t" parameter on the curve where dragging should begin, but don't know how to recalculate control points of that curve.
Here I want to higlight some things related to CorelDraw''s PenTool behaviour that may be used as constaints. I've noticed that when dragging curve strictly vertically, or horizontally, control points of that Bezier curve behave accordingly, i.e. they move on their verticals, or horizontals, respectively.
So, how can I recalculate positions of control points while curve dragging?
Ive just look into Inkspace sources and found such code, may be it help you:
// Magic Bezier Drag Equations follow!
// "weight" describes how the influence of the drag should be distributed
// among the handles; 0 = front handle only, 1 = back handle only.
double weight, t = _t;
if (t <= 1.0 / 6.0) weight = 0;
else if (t <= 0.5) weight = (pow((6 * t - 1) / 2.0, 3)) / 2;
else if (t <= 5.0 / 6.0) weight = (1 - pow((6 * (1-t) - 1) / 2.0, 3)) / 2 + 0.5;
else weight = 1;
Geom::Point delta = new_pos - position();
Geom::Point offset0 = ((1-weight)/(3*t*(1-t)*(1-t))) * delta;
Geom::Point offset1 = (weight/(3*t*t*(1-t))) * delta;
first->front()->move(first->front()->position() + offset0);
second->back()->move(second->back()->position() + offset1);
In you case "first->front()" and "second->back()" would mean two control points
The bezier curve is nothing more then two polynomials: X(t), Y(t).
The cubic one:
x = ax*t^3 + bx*t^2 + cx*t + dx
0 <= t <= 1
y = ay*t^3 + by*t^2 + cy*t + dy
So if you have a curve - you have the poly coefficients. If you move your point and you know it's t parameter - then you can simply recalculate the poly's coefficients - it will be a system of 6 linear equations for coefficients (for each of the point). The system is subdivided per two systems (x and y) and can be solved exactly or using some numerical methods - they are not hard too.
So your task now is to calculate control points of your curve when you know the explicit equation of your curve.
It can be also brought to the linear system. I don't know how to do it for generalized Bezier curve, but it is not hard for cubic or quadric curves.
The cubic curve via control points:
B(t) = (1-t)^3*P0 + 3(1-t)^2*t*P1 + 3(1-t)*t^2*P2 + t^3*P3
Everything you have to do is to produce the standard polynomial form (just open the brackets) and to equate the coefficients. That will provide the final system for control points!
When you clicks on curve, you already know position of current control point. So you can calculate offset X and offset Y from that point to mouse position. In case of mouse move, you would be able to recalculate new control point with help of X/Y offsets.
Sorry for my english