Project 3: Feature Detection

System Implementation Requirements

Determining Mass or MicroCalcification Case and Program Use Cases
Your program will read in a JPEG image (or partial image...as they are very large) from the images in the USF database. Images may or may not have cancer in them (you should test both). Then the user can select a menu option Process->Feature Detect (or alternatively Process->Masses AND Process->Microcalcifications) which will run your blob detecton and 5 different feature detection algorithms.

 

In Either Case the results will be:

• After Blob Detecton will have a Vector (data sructure in java) of Blobs found in the image (for either masses or micro-calcification cases). This vector is calls BlobList and contains instances of the class Blob. This vector will be stored in an instance of the class called CancerAnalysis shown below.
• Based on the size and/or number of blobs you need to decide if you are going to process for case 1 (masses) or case 2 (micro-calcifications). YOU may alternatively have to menu options Process->Mass Detection and Process-Microcalcifications instead of the single Proces->Feature Detect mentioned below.
• CASE 1: Mass Analysis:
  • Blob class (as described more below) will contain minimum of 5 Feature objects f1, f2, f3, f4, and f5 hat are instances of your own implemented sub-classes (Feature1, Feature2, etc) of the class Feature below.
• CASE2: Microcalcification Analysis:
  • depending on what features you propose, you may have features belonging to each blob like in case 1 then:
    • Blob class (as described more below) will contain minimum of 5 Feature objects f1, f2, f3, f4, and f5 hat are instances of your own implemented sub-classes (Feature1, Feature2, etc) of the class Feature below.
  • AND/OR you may have features belonging to the ROI that look at global features between blobs like their relative distribution, density, number, location, etc. In this case you will use the class CancerAnalysis object to not only contain the BlobList but, also these global features cases. These features are called gf1,gf2,gf3,gf4,gf5 and are instances of the classes GlobalFeature1, GlobalFeature2, etc. that are sub-classes of the class Feature you are to create.
• Creation of a n XML file that contains the entire Blob instances in BlobList formatted in XML as described below, called Blobs.xml, that is stored in the same directory as your executable code.
• With the List menu, there are the option List->Blobs. The option List->Blobs will display the nicely formated XML for all of the blobs in a scrollable window.