Find file
Fetching contributors…
Cannot retrieve contributors at this time
26 lines (24 sloc) 1.89 KB
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "">
<html xmlns="">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<title>Untitled Document</title>
<h2>Face Detector</h2>
<p><strong>What You&rsquo;re Seeing:</strong><br />
This face detector draws a red circle around any faces it detects. (There&rsquo;s a size threshold, so distant faces won&rsquo;t be recognized.)</p>
<p><strong>How It Works:</strong><br />
This face detector uses a technique called a <em>cascade classifier</em> to detect faces. The classifier scans across then entire image looking for <em>haar-like features</em> that match those of faces in its training database. It&rsquo;s called a &lsquo;cascade&rsquo; because each candidate region is run through a cascade of different detectors, any of which can reject it. A region is only marked as a facepasses if it passes <em>all </em>of the detectors.</p>
<p><strong>What It&rsquo;s Used For:</strong><br />
Face detection is widely used in from everything to digital cameras to security applications to digital photo albums.</p>
<p><strong>Try and Notice:<br />
</strong>Turn your head to either side, or tilt it slightly. Note that the face detector no longer recognizes your face. That&rsquo;s because this face detector was only trained on full-face images. Other, more sophisticated face recognizers can recognize faces from a wider range of angles.</p>
<p><strong>Learn More:<br />
</strong><a href="">Cascading classifiers</a> on Wikipedia<br />
<a href="">Haar-like features</a> on Wikipedia<br />
<br />
<strong> Credit:<br />
</strong>Based on the OpenCV face detector example.</p>