Assignment -
The purpose of this exercise is to use basic image processing and morphological analysis techniques to quantify reoccurring similar objects in an image. Ideally we want to be able to identify, for example, different cell types and count the number of cells of each type in a microscopy image. However, cells are too complicated for us to work with at the present. Instead, we will sort and count objects that are more tractable-coins.
You will write a Matlab user-defined function that takes as input the image file name as text string, reads in the specified image, uses image intensity and morphological operations for counting, and outputs the numbers of quarters, dimes, nickels and pennies, and total dollar amount, of the coins in the image. For example, one can execute and obtain the following:
>> count_coin('Coins_T1.tiff');
coin count: 7 quarters, 6 dimes, 6 nickels, 8 pennies.
Total amount = $2.73
Specific Instructions:
Work individually without consulting with one another or looking up solutions on the internet. However, you may use Matlab's documentation and help features.
Describe your overall "computational problem" and your computational steps to solve it. Merely listing the Matlab commands used will not be enough.
You may use only the basic Matlab arithmetic and flow control commands and imread, imerode, imdilate, imopen, imclose (including strel) functions, but no other built-in specialty functions from the Signal or Image Processing Toolbox. Violators will forfeit their grade.
(As usual, ask if you are not sure whether a specific function is allowed.)
The body of the code must also be sufficiently and succinctly documented/explained.
Use the provided images to "train" your function (i.e., tweak your settings). You will encounter two types of images, basic and advanced. Basic images (T1 and T2) have isolated coins and coins in contact of neighbors. Advanced images (T3 and T4) in addition contain coins that have small overlap, and one or more foreign coins of the same kind. Your function should work for both types of images, and it will be tested on similar images that are known only to the graders.
Attachment:- Assignment Files.rar