how to identify the cracks from the image

9 次查看（过去 30 天）
ASIM BISAYEE
ASIM BISAYEE 2018-8-24
编辑： Preetham Manjunatha 2025-5-16
  3 个评论
显示 1更早的评论
ASIM BISAYEE
ASIM BISAYEE 2018-8-26
I do have coding in image processing but not to detect the cracks in a segment ways

请先登录，再进行评论。

请先登录，再回答此问题。

回答（2 个）

Image Analyst
Image Analyst 2018-8-26
Try something like a bottom hat filter, imbothat(), then threshold and use regionprops() to thrown out blobs that are vertical. If a slanted crack touches a vertical crack, then you'll have to split them apart with something like watershed.
  10 个评论
显示 8更早的评论
Image Analyst
Image Analyst 2018-8-27
编辑：Image Analyst 2018-8-28
I understand. You're main goal is "trying to develop an algorithm" (programming) rather than material science. Like developing the algorithm is a main part of your Masters thesis or Ph.D. dissertation. So you don't want to buy, or have someone give you, the algorithm because you need to develop it yourself, for your degree. Good luck. Perhaps what I gave you might be a good start.

请先登录，再进行评论。

Preetham Manjunatha
编辑：Preetham Manjunatha 2025-5-16
The image looks quite intricate with regular structures like lines. As @Image Analyst mentioned morphological methods might help to mitigate the non-cracks entities. Here is the MATLAB Crack segmentation and Crack width, length and area estimation codes to calculate/estimate the crack area, width and length. Please try with the morphological crack detection method to get started with. Gradient-based crack segmentation methods can pick the lines heavily in comparision to the morohological approach. Lastly, the semantic segmentation and object detection metrics for the cracks can be found using Cracks binary class bounding box and segmentation metrics package.

类别

在 Help Center 和 File Exchange 中查找有关 Image Processing Toolbox 的更多信息

标签

产品


版本

R2017b

Community Treasure Hunt

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

Start Hunting!

Translated by