O'Reilly logo

Computer Vision with Python 3 by Saurabh Kapur

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

Keypoint localization

In this stage, we will find points that are local extrema. This means, we need to identify points that are best representations of a region of the image (in other words, the neighborhood of the point) in different scales. To locate these keypoints, we iterate over each pixel and compare it with all its neighbors. Now, this is where things start to become interesting. Until now in the book, we always thought of neighbors as the eight pixels that are adjacent to a pixel, but for SIFT, we will not only look at these eight pixels but also at the nine pixels in the preceding images and below this image in the scale space or octave (look at Figure 5). Here we are comparing the pixel value to its 26 neighboring pixels. We select ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required