Image segmentation and measure segmented features

5 ビュー (過去 30 日間)
Abhishek Saini
Abhishek Saini 2021 年 10 月 12 日
コメント済み: Image Analyst 2021 年 10 月 18 日
Hi all,
I want to perform image segmentation as described in the research articles in Fig 5 and 6. I have attached the figure from this article. Figure 59a) and 5(b) are the original images of the microstructure. Now I wan to perform
  1. segmentation as shown in figure 5(c) and 6(c) followed by overlaying the segmented image on originalimage.
  2. Measure te length and with of the segmented features with random line segments by calculating all line intercepts.
Thanks

回答 (1 件)

Image Analyst
Image Analyst 2021 年 10 月 12 日
I didn't read the whole article and can't program it up for you but with just a quick glance it looks like they're segmenting on different hues. So you could convert to HSV color space and get different "maps" for each hue by thresholding the hue channel
hsvImage = rgb2hsv(rgbImage);
hImage = hsvImage(:, :, 1);
color1 = hImage < 0.1;
color2 = hImage >= 0.1 & hImage < 0.2;
color3 = hImage >= 0.2 & hImage < 0.3;
% etc.
If youi want to fill holes of a certain size, or remove blobs less than a certain size, you could also do that with imfill() or bwareaopen().
  8 件のコメント
Image Analyst
Image Analyst 2021 年 10 月 18 日
You can get the dimensions using regionprops()
props = regionprops(mask, 'BoundingBox');
and/or bwferet().

サインインしてコメントする。

Community Treasure Hunt

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

Start Hunting!

Translated by