how to crop a binary image horizontally if vertical length exceeds a certain value?

1 visualización (últimos 30 días)
I have a couple of binary images depicting hand gestures. There are some side effects in those images, that is, images include arm part (down the wrist till elbow). Please view the images attached! I am going to crop this part according to some condition. Is there any way of writing a code for this?
For example: finding the vertical center of white region, or, finding the vertical length of the white region and cut out the bottom part percentage-wise?
Please note that the content of an image (hand) may have different verticality like in the 3rd attached image, so a suggested method should consider up to 40 % curvature as vertical.
Thank you for your ideas!

Respuesta aceptada

KSSV
KSSV el 8 de Mayo de 2017
I = imread('1.jpg') ;
[y,x] = find(I) ;
L = max(x)-min(x) ;
B = max(y)-min(y) ;
Icrop = imcrop(I,[min(x) min(y) L B]) ;
Crops only the region of interest.
  4 comentarios
Sokhib Tukhtayev
Sokhib Tukhtayev el 10 de Mayo de 2017
Sir, in this way I have to manually select the region, but still it is not what I want. It should automatically remove if the image exceeds a certain vertical length. The longer the image vertically is the more part from bottom should be cut.
Thank you anyway!
KSSV
KSSV el 11 de Mayo de 2017
Try this:
I = imread('3.jpg') ;
imshow(I) ;
[y,x] = find(I) ;
x0 = min(x) ; x1 = max(x) ;
y0 = min(y) ; y1 = max(y) ;
%%Form a rectangle
B = x1-x0 ; % horizontal length
L = y1-y0 ; % vertical length
L0 = 200 ; % specified length, change as you wish
if L > L0
dL = L-L0 ;
end
I(x0+dL:end,:) = 0 ;
imshow(I)

Iniciar sesión para comentar.

Más respuestas (0)

Categorías

Más información sobre 3-D Volumetric Image 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!

Translated by