Last time I started a post to provide links to material describing and explaining several popular image processing techniques.
We continue where we left off last time.
Texture segmentation
Texture segmentation is the process of partitioning an image according the the different textures that appear on it. Here is an example
It is a popular processing technique with many applications. Here are a few links that you might find useful.
This is my OpenCV adaptation of a Matlab implementation I found here.
This is segmentation based on Watershed algorithm.
This one is based on Gabor filters and K-means clustering.
Maximally Stable Extremal Regions
It is a method of blob detection. Here is the wikipedia link. In simple words, let’s say you take a picture of a licence plate, a road sign or the writings on a wall and you want to detect where the letters are. You are looking for Maximally Stable Extremal Regions (MSER).
A simple OpenCV example is shown here.
And a Matlab use case here.
Morphological transformations
Simple image shape transformations, but elegant and very useful.
Here is the python OpenCV tutorial.
This is the Matlab imfill functions and this shows how to do it in OpenCV.
These are some more morphological transformations with OpenCV in C++.
An comprehensive explanation of the popular morphological transformations is show here.
Text detection
This is not strictly an image processing technique but more of an image processing task. It is about detecting text in a natural scent and not on scenned documents for example (OCR takes care of that). It can be extended beyond text, to shapes and symbols.
Here is how it is done with Stroke Width Transform, a method backed by a huge industry name.
Another industry giant is behind this technique that uses the Stroke Filter method.
An OpenCV based technique is implemented here.
That was it. I hope this article along with part I, come in handy to you.
I was hoping of writing about Homography and Feature matching, but I guess I will make separate posts about these.
Thanks for reading!
