Hey out there!

1) Reference Images

I decided to use the 300 images provided from the Berkeley Segmentation Dataset and Benchmark as reference images to test my implemented functionality. Results achieved from the MATLAB Image Processing Toolbox are considered to be the “ground truth”.

2) Coding

My first task was to represent a System.Drawing.Bitmap object in a mathematical matrix style. Each element in this matrix holds the gray value of the corresponding pixel in the Bitmap. This was a bit tricky, but finally it works…🙂

I also started with structuring elements which will be needed for morphological operators.  There will be following types:

  •  arbitrary (implemented)
  • square (implemented)
  • rectangle (implemented)
  • diamond
  • line
  • may be some more…

Structuring Element Decomposition is used to decompose large structuring elements in smaller ones to speed up the overall performance of the morphological operator. Decomposition is implemented too for all implemented structuring element types but have not been tested yet.

Finally I need to clean up my code and then I will commit it to the repository…

Welcome to my blog about my work related with the Google Summer of Code 2007 project!

Title Scientific Image Processing Class Library
Student Gernot Margreitner
Mentor Alan McGovern

This project deals with the development of a class library for scientific image processing tasks. The class library should provide a comprehensive set of reference-standard algorithms for tasks like image processing, image analysis or rapid algorithm development in a more scientific approach than existing implementations offer.

The class’ core functionality serves as a basis for scientists and engineers to rapidly build source code based on features like image enhancement, morphological operations, edge detection methods, image segmentation algorithms or filtering techniques.