Machine Vision Research in Wood Industry
Seinäjoki Polytechnic, School of Engineering has done research in wood technology over ten years. The topics have been in the production development projects. 1997 started the Eureka project “European Robotized Network (EuRoboNet)”. In this project we had partners from Sweden, Denmark, Switzerland and Finland. The partners were also Polytechnics or Universities. All partners did their own development projects with companies from their countries. In this EuRoboNet –project we concentrated mostly on use of robots in wood industry to improve the production. Already during this project, there came up the question about possibility to use machine vision systems with robots.
After the EuRoboNet –project ended in 2000 we noticed that we had a good working group and we liked to continue the co-operation we had managed to establish so far. During the earlier project all partners had got questions about automated quality control. So, we developed a new Eureka project “Automated Systems to Recognize Production Results in the Wood Industry (EU-AutoWood)”. In this project we concentrated on the problems in automated quality control and how to use the machine vision systems. The working method in the project was same as before. All partners did projects with theír own company partners and in the meetings we changed experiences and discussed how we had solved occurred problems.
Inspection of the wooden components
During this project Seinäjoki Polytechnic School of Engineering did many kind of development projects for wood industry. The companies were from various sectors of wood industry; doors, log houses, furniture, etc.
For the door factory we made a development project for automating the quality control after the plane. The door components come automatically to the plane where they get planed to the shape. After the shaping, there are not allowed any kind of breaks on the sides of the component or the defects on the component surfaces.
We made the tests with matrix camera in our wood laboratory to inspect the ends of the components. Picture 1. shows the examples of the results.
The surface defects could not be found with matrix camera. It was found out that it wasn’t possible to develop the lighting that all the defects could be found. To find these difficult defects we had the co-operation with one machine vision company and had possibility to use their line scan cameras in trying to solve the problem. The tests went very well and all the defects we wanted to find were found. Picture 2. shows one defect that was found during these tests.
So, all the tests convinced us that the automated quality control in this phase of production can be made by the machine vision system. We also made the financial calculations and asked for the prices of this kind of system. It was found out that the pay back –time for the system is about 2-3 years.
Machine vision in the log house manufacturing
The other example of the Autowood –projects is from the log house industry. The dimensions and machinings were checked with the machine vision system before packing. They were checked from every log. The inspection system got also the information of every log from the bar code attached on it. So it could be noticed immediately if the log was correct or it did not pass the quality control. By this system it was possible to guarantee that all the logs sent to the customer were correct.
The equipment and resources for new projects
During the Autowood –project we provided our wood laboratory with the machine vision cameras, the lighting systems and the conveyors. So, we have good equipment base to make further development projects for the companies working in wood industry. We have also two robots for material handling and can test how to control the robots with the machine vision system.
Seinäjoki Polytechnic is one partner in the new wood automation project called TUPAS, financed by Tekes. The target of this project is to develop wood industry production by means of automation. The companies get help for their development projects from the partners that are for example polytechnics. Part of the financing comes from Tekes. The development projects can handle robotizing, quality control including machine vision, production simulation and all other methods for developing the production by automation.