<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Top-Hat on Aayush Bajaj's Augmenting Infrastructure</title><link>https://abaj.ai/tags/top-hat/</link><description>Recent content in Top-Hat on Aayush Bajaj's Augmenting Infrastructure</description><generator>Hugo</generator><language>en</language><copyright>© 2026 Aayush Bajaj</copyright><lastBuildDate>Thu, 12 Nov 2026 02:04:46 +1100</lastBuildDate><atom:link href="https://abaj.ai/tags/top-hat/index.xml" rel="self" type="application/rss+xml"/><item><title>Morphology: Counting Nuts and Bolts</title><link>https://abaj.ai/wiki/computer-vision/morphology/</link><pubDate>Wed, 11 Nov 2026 21:00:00 +1100</pubDate><guid>https://abaj.ai/wiki/computer-vision/morphology/</guid><description>&lt;p>A threshold almost never hands you a clean mask. You get specks of noise in the background, pinholes and cracks in the objects, ragged edges, objects cut off by the frame and objects that touch. &lt;em>Mathematical morphology&lt;/em> is the toolkit for repairing such masks. It treats a binary image as a set of pixels and probes it with a small set of offsets. Everything is built from two primitives, one that grows objects and one that shrinks them, together with plain set algebra. There are no derivatives and no weights: the operations are nonlinear, and they judge structures by &lt;em>size and shape&lt;/em> rather than by contrast.&lt;/p></description></item></channel></rss>