No Title

Tue, 30 Nov 1999 00:11:00 EST

Krzysztof,

We've not done a project like that before so you would have to find someone to help you understand those papers and implement a similar technique. Please keep in mind that research projects are typically done in ideal settings so quite a bit of work is needed in order to turn research results into a reliable and workable solution.

At least the links you mention will provide some examples. You will need to create a large database (at least 500 samples) in order to know that the techniques work for you. Because your lighting and types of wood might differ from the research papers, you may find that other techniques work better. A lot of experimental work is required.

STeven.

Wood defects detection

Tue, 30 Nov 1999 00:11:00 EST

Dear Sirs,

I would like to ask is there any chance to detect wood defects (especially knots type) using RoboRealm?

I would like to achieve something like here:

https://www.researchgate.net/publication/220872249_Lecture_Notes_in_Computer_Science

https://www.researchgate.net/publication/220665710_Automated_visual_inspection_system_for_wood_defect_classification_using_computational_intelligence_techniques

http://www.ee.oulu.fi/research/imag/knots/KNOTS/

We are in construction of automatic defect(especially decayed knots) repairing machine and I'm looking for solution for detection. In first prototype detection is done by people who mark with stamp knots to repair. Then this stamp is recognized using "shape recognition" by RoboRealm - it works quite well.

But I'm thinking to go further and detects decayed knots with the software only but the knots shape is not consistent and hard to evaluate which knot is ok or not ok.

Any help?

Best regards,
Krzysztof

RoboRealm Forum

No Title

Wood defects detection