How do 3D coordinates map to a 3D matrix?
9 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
ting
el 21 de Nov. de 2013
Comentada: Bruno Pop-Stefanov
el 22 de Nov. de 2013
The project is about point cloud. Given a large number of 3-D coordinates (x-y-z) with all intensities are the same, say 1 or 255.
The range of x, y and z shown as follows.
* -113<x<158;
* -136<y<135 and
* -771<z<-500.
# If i create a 272 by 272 by 272 matrix, how can i map those 3-D coordinates to the matrix?
A = zeros(272, 272, 272);
# I know the matrix index starts from 1 in MAT-LAB. Should I create a 3-D grid to do mapping?
3 comentarios
Respuesta aceptada
Bruno Pop-Stefanov
el 21 de Nov. de 2013
I assume in the following piece of code that your 3D points are listed in an N-by-3 vector named 'points'.
% Min and max values
xmin = -113;
xmax = 158;
ymin = -136;
ymax = 135;
zmin = -771;
zmax = -500;
% Initialize output matrix in which a non-zero entry indicates a 3D point exists in the input set
mat = zeros(xmax-xmin+1, ymax-ymin+1, zmax-zmin+1);
% For all 3D points
for p=1:size(points, 1)
if ( points(p,1)>xmax || points(p,1)<xmin || points(p,2)>ymax || points(p,2)<ymin || points(p,3)>zmax || points(p,3)<zmin )
fprintf('Error: point %d is out of bounds.\n', p);
else
i = points(p,1) - xmin + 1;
j = points(p,2) - ymin + 1;
k = points(p,3) - zmin + 1;
mat(i, j, k) = mat(i, j, k) + 1;
end
end
The output matrix 'mat' contains non-zero entries when a 3D point exists, but you'll have to convert into indices. For example, if the first point in 'points' is [29 -76 -614], then
mat(29-xmin, -76-ymin, -614-zmin) ~= 0
2 comentarios
Bruno Pop-Stefanov
el 22 de Nov. de 2013
If all your data points in your input vector are unique, then, yes, the matrix will only have 0's and 1's. Coordinate (i,j,k) in the output matrix is the number of points in the input vector that have corresponding coordinates (x,y,z). If you only want 0's and 1's, you can change line
mat(i, j, k) = mat(i, j, k) + 1;
to
mat(i, j, k) = 1;
Yes, (xmin, ymin, zmin) maps to (1, 1, 1).
Más respuestas (0)
Ver también
Categorías
Más información sobre Point Cloud Processing en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!