why threholded image didn't show objects

Question

Tian Tian 2017-8-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/353162-why-threholded-image-didn-t-show-objects

评论： Image Analyst 2017-8-19

Hi, I have a binary image (multithreholded; 3472 * 3256), and now I want to get objects larger than 50 pixels in area, but every time the output is nothing (pure white or pure black when i changed 50 to others). Does it mean it regards the whole as one object after threholding? Mine is like grain distributed image. Could anyone help me? Thank you.

3 个评论
显示 1更早的评论隐藏 1更早的评论

Tian Tian 2017-8-19

untitled.fig

The attached is the figure outputted by matlab after threholding (I used multithrehold, then converted to binary). I want to get the grain size distribution (e.g.how many objects have area of 50 pixels), shape (how many objects have cicurlarity of some degree), etc. Could you help me？ Many thanks.

Image Analyst 2017-8-19

See my Image Segmentation Tutorial

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Image Analyst 2017-8-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/353162-why-threholded-image-didn-t-show-objects#answer_278425

在 MATLAB Online 中打开

Use either bwareafilt() or bwareaopen()

binaryImage = bwareaopen(binaryImage, 50);

or

binaryImage = bwareafilt(binaryImage, [50, inf]);

If you have a multilevel image, like 3 gray levels, then extract just one of the numbers to form a binary image:

binaryImage = grayImage == 1; % or whatever number you want.

2 个评论
显示无隐藏无

Tian Tian 2017-8-18

Many thanks! Just a quick question, after multithresholding, the image is in black, gray and white (I only want white), then I found binaryImage = grayImage == 3 can help to extract white out, does this make sense?