Face Detection via Skin Color Segmentation

MATLAB 3K 230 views 0 downloads 1 credits
Tags:

Resource Overview

Face detection using skin color segmentation involves converting color images from RGB to YCbCr color space, segmenting skin regions through threshold operators, obtaining binary images, followed by dilation and erosion operations, and finally performing edge detection.

Detailed Documentation

Face detection is an image processing technique used to segment skin color regions in color images. This technique involves converting color images from RGB color space to YCbCr color space (using conversion formulas like Y = 0.299R + 0.587G + 0.114B, Cb = 128 - 0.169R - 0.331G + 0.500B, Cr = 128 + 0.500R - 0.419G - 0.081B) and applying threshold operators (typically with Cr values between 133-173 and Cb values between 77-127) to segment skin regions, resulting in a binary image. Subsequently, morphological operations such as dilation (expanding skin regions using structuring elements) and erosion (refining boundaries by removing small artifacts) are applied to further extract clean skin regions. Finally, edge detection algorithms (like Canny or Sobel operators) can be implemented to capture the contour edges of facial features through gradient calculation and non-maximum suppression.

You May Also Like