Igor Polk Image Analysis Lab

Please, contact me at ipolk@virtuar.com , Home | Blog

Manifesto

I am researching image analysis trying to find the way image processing and recognition may be done in nature by Earth organisms like us. Presumably, these processes are parallel and simple. As a unit, I have chosen a notion of "a dale", which someone might call "concavity". The term name "dale" better suits gray images, the images I am working with. My hypothesis is that we recognize these "dales", classify and build hierarchy of them and this represents objects. The goal is to show the robustness and applicability of methods based on this idea as well as to show that it is one of the methods used by living organisms in image recognition. So far the results are promising, but a lot of research is required.

News

2017-01-31
Results of research:
1. "Interaction" of objects in the image by tracking how minimal changes on one object affect other objects.
2. Method of image superposition to generate dales over images without dales.
Link

2016-10-06
New article on software engineering "Unified Control and Data Flow Diagrams Applied to Software Engineering and other Systems"(pdf) on arXiv.org.

2016-08-28
"Asynchwave" image analysis c++ template library is published at: https://sourceforge.net/projects/asynchwave/. This is a research project.
An article " Construction of Convex Sets on Quadrilateral Ordered Tiles or Graphs with Propagation Neighborhood Operations. Dales, Concavity Structures. Application to Gray Image Analysis of Human-Readable Shapes" is ready and published on arXiv.org: https://arxiv.org/abs/1608.08251.

Read more about it and watch animated images

What is a tiling?

2013-03-19

A"Asynchronous Cellular Operations on Gray Images Extracting Topographic Shape Features and Their Relations" is submitted to arxiv.org on March 19, 2013.

Availability of code is under consideration. Anyway, it has to be cleaned up a little to be published.

Sept 1, 2013
Clean-up of the code is well under way..

Keywords

Symbol, character, hieroglyph form classification, mathematical morphology, cellular automata, graph theory.

My Groups, Communities

Artificial Intelligence, TRIZ, and systems science
Patterns, mathematics, and science

Publications

"A memory to read out and write a block of words via one memory access and its application to increase processor and cache performance", 1992 - link

Mathematics

The reason I'd like to publish articles on this topic of elementary math is that however sadly is sounds, we, engineers, do not know math. Just plainly we do not. We may be taught and we may even remember "how to use some of mathematical results" but the math is not the same like we have learned it at schools. But we must not learn math the same way as mathematicians do. We have different goals. We must find a way to present and use the real math in our professional life. I mean "math effects"... After getting acquainted a little with "the real math", sadly, I came to conclusion that we should start from the very beginning. In this way:

On Euclid Proposition 5, Book 1

 

Home


Links

http://www.absoluteastronomy.com/topics/Mathematical_morphology

Test charts for videcameras ( on the image-engineering website )

Not related:

"An essay on patterns in musical composition transformations, mathematical groups " + fibonacci, group theory