Lecture

For an image, it consists of two parts:

  1. Background
  2. Foreground

A Structured Element (SE) has three parameters to consider:

  1. How much
  2. Different Shapes
  3. Different Size

NOTE: Look back at structuring elements and the use of erosion. Example: if you have a 5 x 20 image, where 5 is the width and 20 is the height, if you want to remove the amount of pixels vertically, you do d/2; if you want to remove the amount of pixels horizontally, you do d/4.

+—+ | 1 | <- removes one pixel at the top +—+ +—+ | 1 | <- this is the origin +—+ +—+ | 1 | <- removes one pixel at the bottom. +—+

The above graphic is used for erosion.

Opening and Closing

  • Opening - It smoothens the contour of an object. Is ideal for eliminating small objects. Erosion followed by dilation.
  • Closing -

MatLab: Morphology, Opening, Erosion, Dilation

In MatLab, in order to use morphology, you use the function: strel();

To use an Opening, you use imopen();

To use erosion, you use imerode();

To use dilationm you use imdilate();

Hit-or-miss Transform

It is a basic tool for shape detection in binary images. Performs erosion on the image w/ first structuring element and on the complement of the image w/ the second structuring element.

Reading

Summary