how to crop white regions alone in a binary image without user interface.

Question

Ashwini 2016-2-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/266179-how-to-crop-white-regions-alone-in-a-binary-image-without-user-interface

评论： shan shan 2020-3-12

I have a binary image with background in black color and foreground in white color. how to crop only the foreground region without using tools like roipoly or imfreehand..

to

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

shan shan 2020-3-12

This is my binary image from this i have to crop the white regions into three images separately can you please help me to do this?,thanks in advance.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Image Analyst 2016-2-2

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/266179-how-to-crop-white-regions-alone-in-a-binary-image-without-user-interface#answer_208199

在 MATLAB Online 中打开

You can use regionprops() to get the bounding box. Then use indexing to extract (crop) it.

Alternatively, if you don't have the Image Processing Toolbox, use

[rows, columns] = find(binaryImage);
row1 = min(rows);
row2 = max(rows);
col1 = min(columns);
col2 = max(columns);
croppedImage = binaryImage(row1:row2, col1:col2);

2 个评论
显示无隐藏无

Ashwini 2016-2-12

在 MATLAB Online 中打开

Thank a lot.I executed the above code and it works.. But how to get this with regionprops() ?? when i use regionprops(img,'Boundingbox'), boundingbox gives fraction values. i tried with a binary matrix z=

   0     0     1     1
   0     0     1     0
   0     0     1     0
   0     0     0     0
   0     1     1     1

st=regionprops(z,'boundingbox'); gives 1x1 struct which has Boundingbox=[0.5000 0.5000 5 5]

how to get index from this.. can u provide me the code...

Image Analyst 2016-2-12

There is nothing to crop. You want the whole image. regionprops gives box coordinates "outside" the pixels because the pixels are at the integers for example if a pixel at 3 was zero and the pixel at 4 was 1, then the bounding box would go in between them at a location of 3.5.

请先登录，再进行评论。