<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Güven YILMAZ on Medium]]></title>
        <description><![CDATA[Stories by Güven YILMAZ on Medium]]></description>
        <link>https://medium.com/@yilmazguven247?source=rss-09de0e74d2df------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*zCKJ9n90P90ntfI6</url>
            <title>Stories by Güven YILMAZ on Medium</title>
            <link>https://medium.com/@yilmazguven247?source=rss-09de0e74d2df------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 17:38:08 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@yilmazguven247/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Convolution in Image Processing: Teaching Computers How to See]]></title>
            <link>https://medium.com/@yilmazguven247/convolution-in-image-processing-teaching-computers-how-to-see-eecff3d9aec8?source=rss-09de0e74d2df------2</link>
            <guid isPermaLink="false">https://medium.com/p/eecff3d9aec8</guid>
            <category><![CDATA[image]]></category>
            <category><![CDATA[image-processing]]></category>
            <category><![CDATA[convolution]]></category>
            <dc:creator><![CDATA[Güven YILMAZ]]></dc:creator>
            <pubDate>Thu, 21 Aug 2025 13:48:59 GMT</pubDate>
            <atom:updated>2025-08-21T13:48:59.931Z</atom:updated>
            <content:encoded><![CDATA[<p>Imagine this: you take a photo, and the camera instantly recognizes your face. Or you open Instagram, tap on a filter, and suddenly your skin looks flawless. Behind all of this lies one mission: teaching computers how to see. One of the most powerful tools in that mission is convolution. It’s like a hidden trick that allows computers to highlight edges, blur images, sharpen details, or bring out specific patterns. Today, we’ll take a beginner-friendly look at convolution. Don’t worry — no endless formulas. Instead, we’ll use intuition, analogies, and small examples to make things crystal clear.</p><h3><strong>1. An Image = Painting with Numbers</strong></h3><p>Here’s a key truth: computers don’t see images the way we do. For them, every photo is just a giant matrix of numbers.</p><p>Grayscale image → each pixel is a number between 0 and 255.</p><p>0 = black 255 = white In between = shades of gray</p><p>Color image → each pixel is made of 3 numbers: Red (R), Green (G), Blue (B). Sometimes its BGR but you get the point.</p><p>So, while you admire a landscape photo, your computer only sees millions of tiny numbers in neat grids.</p><h3><strong>2. What Is Convolution?</strong></h3><p>Think of convolution like holding a magnifying glass with a special pattern carved into it, then sliding it across a picture. If the magnifying glass highlights edges, you suddenly see outlines pop. If it’s designed for blur, the world looks softer, as if you’ve taken off your glasses. If it’s for noise reduction, distracting details quietly disappear. In short: convolution means sliding a small filter across the image and creating a brand-new version of it.</p><h3><strong>3. How Does It Work?</strong></h3><p>Let’s play a quick game. We have this tiny image matrix:</p><pre>[[1, 2, 3],<br> [4, 5, 6],<br> [7, 8, 9]]</pre><p>And here’s our filter:</p><pre>[[0, 1, 0],<br> [1,-4, 1],<br> [0, 1, 0]]</pre><p>We place the filter at the center. Multiply corresponding numbers, add them up… Result? → 0 That number is written into a new matrix. As we slide the filter across the entire image, pixel by pixel, a transformed picture emerges. What changed? Now certain features — edges, textures, or shapes — stand out more clearly than before.</p><h3><strong>4. Two Magic Helpers: Padding and Stride</strong></h3><p>Convolution has two little assistants that control how it behaves:</p><p>Padding: Imagine putting a border of zeros around the image so the filter won’t “fall off” at the edges.</p><p>Stride: How many pixels the filter moves each step.</p><p>Stride = 1 → move one pixel at a time → very detailed. Stride = 2 → skip ahead → faster, but less detail.</p><p>Think of it like walking: Small steps → you notice everything. Big steps → you move quickly but miss subtle details.</p><h3><strong>5. Where Do We Use It?</strong></h3><p>Everywhere. Convolution powers: — Face recognition (your phone unlocking by looking at you) — Filters (blur, sharpen, smooth) — Noise reduction (restoring old photos) — Deep learning (CNNs extracting features like “cat ears” or “car wheels”)</p><p>If modern image processing is a toolbox, convolution is the Swiss Army knife.</p><h3><strong>6. The Bigger Picture</strong></h3><p>Mathematically, convolution looks like this:</p><pre>O[i,j] = ∑∑ I[i+m, j+n] * K[m,n]m</pre><p>Where: I = input image K = filter (kernel) O = output image</p><p>But here’s the part that truly matters: Convolution helps computers decide what in an image is important.</p><p><strong>Conclusion</strong></p><p>Today, you’ve seen convolution not just as a formula but as a way for computers to “see.” A filter slides across an image, pixel by pixel, and creates a new version that highlights what matters most.</p><p>The real magic? In Convolutional Neural Networks (CNNs), you don’t even tell the computer which filters to use. The network learns them on its own. It discovers edges, shapes, and patterns — just like our brains do when we look at the world.</p><p>Convolution isn’t just math. It’s the first step in teaching machines how to interpret the visual world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eecff3d9aec8" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>