how to connect the adjacent lower edges of each bounding box

1 visualización (últimos 30 días)
praveen rai
praveen rai el 14 de Ag. de 2019
Editada: praveen rai el 17 de Ag. de 2019
I have drawn a bounding box of rectangle shape on each character and save the all 4 point of all bounding box
Now I want to connect the lower edges of bounding box with each other
How to do it??
sample code for making bounding box and saving all 4 points of bounding box
I = imread('textlineImage.png');
I=im2bw(I);
I=imcomplement(I);
st=regionprops(I,'BoundingBox','Area');
allBB = zeros(length(st), 4);
for k = 1 : length(st)
thisBB = st(k).BoundingBox;
rectangle('Position', [thisBB(1),thisBB(2),thisBB(3),thisBB(4)],...
'EdgeColor', 'r', 'LineWidth', 1)
allBB(k, :) = thisBB;
points are save in varibale allBB
Image is
234.jpg
  8 comentarios
Rik
Rik el 17 de Ag. de 2019
You have the coordinates right? Then what exactly is your issue?
praveen rai
praveen rai el 17 de Ag. de 2019
ya I have 4 point values of rectangle box in variable allBB but which two points I have to take and how to plot for every line so that I can get spline that is the issue
I am unable to wirte these thing in loop

Iniciar sesión para comentar.

Respuestas (1)

Image Analyst
Image Analyst el 17 de Ag. de 2019
Editada: Image Analyst el 17 de Ag. de 2019
Use kmeans on the y values to identify boxes in one of the 5 lines that your image has.
[classIndexes, classCentroid] = kmeans(y, 5);
After that, you know which line each y is in then sort by x, then use spline. Something like (untested)
y = allBB(2:4:end);
[classIndexes, classCentroid] = kmeans(y, 5);
newXSamplePoints = 1 : size(yourImage, 2)
for k = 1 : 5
thisClass = classIndexes(k);
thesex = sort(x(thisClass), 'ascend');
% Fit the x to splines at every column in the image.
newY = spline(x, y, newXSamplePoints);
end
hold on;
plot(newXSamplePoints, newY, 'm-', 'LineWidth', 2)
  1 comentario
praveen rai
praveen rai el 17 de Ag. de 2019
Editada: praveen rai el 17 de Ag. de 2019
its showing the following error
Error using kmeans (line 247)
X must have more rows than the number of clusters.
Error in rawcode (line 25)
[classIndexes, classCentroid] = kmeans(y, 5);
allBB has value 208*4 double and y has 1*208 double

Iniciar sesión para comentar.

Etiquetas

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by