5.Image_segmentation

# The functions edge 2 through edge 4 compute a similar

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: through Edge 4 compute a similar type of one dimensional edge property. The operators as shown above also incorporate some image smoothing, since they accumulate two measurements of edge contrast and then add them together. This type of operator will diminish the effect of an errant intensity point. Sobel operator The Sobel operator incorporates edge information in two directions as follows: Edge = (<A+2B + C - G – 2H – I >2 + < A + 2D + G – C – 2F – I>2) ½ This operator computes a weighted average intensity function along the borders of the subimage, and then forms two edge measurements perpendicular to each other. The two edge properties are then combined in a “quadrature” measurement. The Sobel operator generally enhances edges in an acceptable fashion. The squaring and square root operations will be very time consuming, so one normally would modify this operator by replacing the squares with absolute values and eliminating the square root. This is referred to as the modified Sobel operator. Contrast Operator Another approach to edge detection is the use of contrast differences. Consider the following: Edge = E – (A + B + C + D + F + G + H + I )/ 8 This operator compares the intensity of the central pixel to the pixel’s surrounding. Here, only t he difference in contrast between the central pixel and its neighbors is considered, regardless of the distribution of the neighboring pixel intensities. Edge-based Segmentation Edge-based segmentation represents a large group of methods based on information about edges in the image, it is one of the earliest segmentation approaches. Edge-based segmentations rely on edges found in an image by edge detecting operators -- these edges mark image locations of discontinuities in gray level, color, texture, etc. Image resulting from edge detection cannot be used as a segmentation result. Supplementary processing steps must follow to combine edges into edge chains that correspond better with borders in the image. The final aim is to reach at least a partial segmentation -- that is, to group local edges into an image where only edge chains with a correspondence to existing objects or image...
View Full Document

## This document was uploaded on 03/12/2014 for the course MECHANICAL 214 at University of Manchester.

Ask a homework question - tutors are online