Göm meny

Automatic identification of wooden panels

In today’s highly automated manufacturing plants for producing wooden products, it is useful to be able to track individual objects through the entire production chain. Existing marking methods are either expensive or associated various technical issues. Since every individual wooden surface has a unique growth ring pattern, a camera should be able to recognise objects without the use of added markings. The aim of the thesis is to select and evaluate a method for identification of individual wooden surfaces, with assistance from OptoNova’s algorithm development team.

OptoNova possesses a large amount of annotated image material, tools for development and evaluation of image processing and machine learning algorithms, and a laboratory where a prototype system consisting of cameras and transporters can be built. Experts in algorithm development and optical physics will be able to supervise and provide feedback during the project. A successful project may lead to deployment in large manufacturing industries and/or a published article.

This thesis is suitable for students interested in computer vision and/or machine learning. Even though the project is focused primarily on method development, programming experience is necessary to work effectively with existing frameworks.

Contact

Josef Grahn


Senast uppdaterad: 2014-03-18