清空筛选条件
清空筛选条件

In binary image, how to convert large black pixels area into white ?

4 次查看（过去 30 天）
Y.L.K KHUMAN
Y.L.K KHUMAN 2017-12-7
评论： Y.L.K KHUMAN 2017-12-9
I am working on skew detection and correction. I had corrected the skewness of the image, but I found that there is large size of Black areas on all sides of the corrected image. I tried to convert the black pixels (large area only) to white using the following codes. But it is not working properly. Sometimes its also remove black pixel from the pictures itself and sometimes it does not remove the larger area of the black pixels.
<<
>>
measurements = regionprops(SkewCorrectedImage == 0, 'Area','PixelIdxList');
Regions = find([measurements.Area] >=200000);%%
theIndexes = vertcat(measurements(Regions).PixelIdxList);
c=SkewCorrectedImage;
(theIndexes)=1
  2 个评论
Jan
Jan 2017-12-7
You should be able to set the added pixels to white directly in the code for rotating the image. This would be much smarter that using an auto-detection of the black triangles afterwards. So please post the code to obtain the shown image.
Y.L.K KHUMAN
Y.L.K KHUMAN 2017-12-7
编辑：Y.L.K KHUMAN 2017-12-7
Here is the rotation part code:
props = regionprops(joinedBlobs, 'BoundingBox','Orientation'); % Find bounding box.
boundingBox = props.BoundingBox;
angle = props.Orientation
angleToRotate = -angle(1);
figure, imshow(IPImg), rectangle('Position', boundingBox, 'EdgeColor', 'red');
croppedImage = imcrop(IPImg, props.BoundingBox);
figure, imshow(croppedImage), title('cropped Image');
SkewCorrectedImage = imrotate(croppedImage, angleToRotate);
figure, imshow(SkewCorrectedImage), title('corected image');
imsave();

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

Akira Agata
Akira Agata 2017-12-7
You can easily do that by using imclearborder function, like:
BWout = ~yourBinaryImage;
BWout = imclearborder(BWout);
BWout = ~BWout;
  1 个评论
Y.L.K KHUMAN
Y.L.K KHUMAN 2017-12-9

Thank for your help. When I try this code, it removes some boundary part, which is also a part of the image. Rotated Image:

After Using your code:

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Geometric Transformation and Image Registration 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by