Question about blob detection matlab algorithm
显示 更早的评论
|Hello every body, am newbie in Matlab and am actually interested in blob detection mecanism:
check the link below :
https://fr.mathworks.com/help/vision/deep-learning-semantic-segmentation-and-detection.html
it is said that:
*Blob analysis uses segmentation and blob properties to identify objects of interest*
can any body explain to me how this algorithm works (segmentation) and what blob properties are used?
any help will be highly appreciated |
采纳的回答
Image Analyst
2018-10-14
2 个投票
3 个评论
djim djim
2018-10-17
Image Analyst
2018-10-17
1. Segmentation. First you segment the image to get a binary image, for example you could threshold your image at some brightness level. Or you could go through a lot of different steps but eventually you get to a point where you have an image that can be thresholded and produce a binary image.
2. Connected components labeling. Then it does a connected components analysis on the binary image to determine which "true" pixels are connected and belong to the same blob. In this way, every connected blob has its own ID label that uniquely identifies it. See this Wikipedia link for more explanation.
3. Analysis/measurement. Then it goes through a variety of algorithms to identify area, perimeter, or whatever you asked for. The exact algorithm depends on what measurement is being made.
djim djim
2018-10-18
更多回答（0 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Computer Vision with Simulink 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
