Building Machine Learning Systems with Python - Richert, Coelho

Computer Vision – Pattern Recognition
Finally, we can combine the two images to have the center in sharp focus and the
edges softer.
ringed = mh.stretch(im*C + (1-C)*im12)
Now that you know some of the basic techniques of filtering images, you can build
upon this to generate new filters. It is more of an art than a science after this point.
Pattern recognition
When classifying images, we start with a large rectangular array of numbers (pixel
values). Nowadays, millions of pixels are common.
We could try to feed all these numbers as features into the learning algorithm. This is
not a very good idea. This is because the relationship of each pixel (or even each small
group of pixels) to the final result is very indirect. Instead, a traditional approach is to
compute features from the image and use those features for classification.
There are a few methods that do work directly from the pixel values. They have
feature computation submodules inside them. They may even attempt to learn
what good features are automatically. These are the topics of current research.
[ 210 ]