# Team:Aachen/Errorlog

(Difference between revisions)
 Revision as of 09:45, 19 June 2014 (view source)Mjoppich (Talk | contribs)← Older edit Revision as of 19:12, 21 June 2014 (view source)Mjoppich (Talk | contribs) Newer edit → Line 65: Line 65:

- The third part of our project is our software Measurarty. The software will allow us to analyze the fluorescence emitted by our sensor cells in a more advanced and better way than just using a simple Treshold. We will utilize a modern segmentation algorithm in combination with further, detailed image processing algorithms. + The third part of our project is our software Measurarty. The software will allow us to analyze the fluorescence emitted by our sensor cells in a more advanced and better way than just using a simple tresholding technique. We will utilize a modern segmentation algorithm in combination with further, detailed image processing algorithms.

+

- More information will be available soon, as soon as we can describe the theoretical foundations of the software in detail. But therefore we'd love to use formula that need to be displayed properly using the math-tag :-) + At the core of our segmentation algorithm is a Statistical Region Merging1 implementation written for Qt enabled C++. -
+ Statistical Region Merging (SRM) is a straight-forward technique for creating regions in an image. Such regions then can be used for region-based image segmentation, the core of most image analysis softwares. Additionally, segmenting regions, instead of single pixels, brings an enormous performance improvement. -
+ To the knowledge of this team, most SRM implementations only use region-size and color gradients for region creation. - While MathJax seems to work $$2+2 \cdot \frac{1}{2}$$,
+ We are going to change that. - the math-tag 2 \cdot \frac{1}{2} does not ... +

-

+

Over the next few months the computational devision of our team will develop an easy-to-use, easy-to-run software for image analysis of color-based image analytics.
- At the core of our segmentation algorithm is a Statistical Region Merging1 implementation written in Qt enabled C++. +
+ For this purpose we would like to present our novel approach in this wiki, and formulas will help you to understand our algorithm. Until the math-tag in MediaWiki is working again, MathJax will provide us with a $\LaTeX$ environment!

# Device and Software

The soft- and hardware aspects of our project are described in the Error Log, because we all learn from errors we made.

# The Software: Measurarty

The third part of our project is our software Measurarty. The software will allow us to analyze the fluorescence emitted by our sensor cells in a more advanced and better way than just using a simple tresholding technique. We will utilize a modern segmentation algorithm in combination with further, detailed image processing algorithms.

At the core of our segmentation algorithm is a Statistical Region Merging1 implementation written for Qt enabled C++. Statistical Region Merging (SRM) is a straight-forward technique for creating regions in an image. Such regions then can be used for region-based image segmentation, the core of most image analysis softwares. Additionally, segmenting regions, instead of single pixels, brings an enormous performance improvement. To the knowledge of this team, most SRM implementations only use region-size and color gradients for region creation. We are going to change that.

Over the next few months the computational devision of our team will develop an easy-to-use, easy-to-run software for image analysis of color-based image analytics.

For this purpose we would like to present our novel approach in this wiki, and formulas will help you to understand our algorithm. Until the math-tag in MediaWiki is working again, MathJax will provide us with a $\LaTeX$ environment!

1. [Nock, R.; Nielsen, F., "Statistical region merging," Pattern Analysis and Machine Intelligence, IEEE Transactions on , vol.26, no.11, pp.1452,1458, Nov. 2004]

# The Device: Cellock Holmes

You may wonder: again Cellock Holmes?! YES!
In lack of a better name, our name for the device is Cellock Holmes, again.

# The Other projects: Arbitrary Stuff

We're doing some collaborations and side projects for iGEM as well. We'll inform you about that here :)