Image enhancement to extract key features

4 次查看（过去 30 天）
Rohan Kokate
Rohan Kokate 2023-8-24
回答： Balaji 2023-9-4
Hello,
I have high-speed images of flow inside a channel (attached Fig. 1). I want to extract some key features like the interface of the bubble and the bubble nucleation sites on the wall. The walls of the channel (the gray portion of the image) is not of importance for me. So I tried to invert the map of the image and then try to adjust the brightness and contrast manually and it looks something like Fig. 2.
This approach, in certain cases, results in distorting the bubbles on the wall and only captures the long bubble slugs in the channel. I am new to image processing so I am not really sure about the best way to address this particular issue. How should I go ahead with this? I am sure there is a much betetr way to do this as I'm essentially trying to brute force what I want. Also what would be the best way to automate this analysis for a bunch of images?
Thanks in advance.
Rohan.
  2 个评论
Matt J
Matt J 2023-8-24
In addition ot the raw image, please provide a marked-up image showing the things you want extracted.
Rohan Kokate
Rohan Kokate 2023-8-24
Apologies! I have attached a picture showing the features I want to extract along with the raw images of the same in the zip file.

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

Balaji
Balaji 2023-9-4
Hi Rohan,
You can try the following steps to extract the features from your image.
Thresholding: Instead of manually adjusting the brightness and contrast, you can use thresholding techniques such as Global Thresholding, adaptive thresholding, or Otsu’s thresholding to get best results for your image
https://www.mathworks.com/help/images/ref/graythresh.html
https://www.mathworks.com/help/images/ref/multithresh.html
Morphological Operations: Morphological operations such as Erosion and dilation methods can be used to smooth bubble edges, eliminate noise, and fill in gaps.
https://www.mathworks.com/help/images/morphological-filtering.html
3. Edge detection: To extract the bubble interface, you can use edge detection algorithms such as the Canny edge detector or the Sobel operator.
https://www.mathworks.com/help/images/edge-detection.html
4. Blob analysis: Once you have the segmented bubbles, you can perform blob analysis to identify and measure properties of the bubbles, such as their size, shape, and position.
https://www.mathworks.com/help/images/ref/regionprops.html
Hope this helps!
Thanks
Balaji

类别

在 Help Center 和 File Exchange 中查找有关 Image Segmentation and Analysis 的更多信息

标签

产品


版本

R2022b

Community Treasure Hunt

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

Start Hunting!

Translated by