Illumination design of a microrobotic platform
Hänninen, Antti Johannes (2013)
2013
Automaatiotekniikan koulutusohjelma
Teknisten tieteiden tiedekunta - Faculty of Engineering Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2013-05-08
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201305171136
https://urn.fi/URN:NBN:fi:tty-201305171136
Tiivistelmä
This thesis discusses the optimization researchof on illumination design in of a microrobotic platform. The platform is built to automatically perform individual paper fiber (IPF) and fiber bond measurements of a single bond in order to harvest detailed information about paper qualityunderstanding on paper strength. Currently, mathematical and exist to estimate the microscale properties of pulp and paper. Mechanical and chemical properties of paper fibers are usually determined by bulk parameters or indirectly. Some fiber parameters have been characterized by using current tools, but new equipment is needed for higher yields and fiber-fiber bond measurements. Detailed individual fiber level measurement will provide important knowledge of individual paper fibers and bonds. This platform will also allow the possibility to perform and examine chemical manipulations on a single fiber or bond to improve the qualitystrength properties of paper.
The platform is searching for the fibers using two cameras for a 3D-view of the fibers. It is equipped with two microgripper actuators to perform the grasping of IPF or two bonded fibers. The main task for this thesis was to improve the image quality by optimizing the fiber illumination for the two cameras. Secondary tasks were to check that this solution will also work in finding the actuators from the image and possibly allowing the bonded area measurements visually.
Evaluating the suitable illumination was performed by testing different sorts types of lighting illumination architectures and improving the most successful methods further. The first candidates were compared taking a set of sample images and calculating the fiber recognition performance for each.
Finally a new illumination module design was created designed to allow the use of special polarized backlight illumination inside the platform structure previously designed. This polarized illumination works using two crossed polarization filters behind and in front of the fibers to block direct light but letting through the light refracting and reflecting from the fibers thereby changing the type or direction of its polarization. The resulting image then has a good contrast between the black or dark background and the white or light coloured fiber targets.
The platform is searching for the fibers using two cameras for a 3D-view of the fibers. It is equipped with two microgripper actuators to perform the grasping of IPF or two bonded fibers. The main task for this thesis was to improve the image quality by optimizing the fiber illumination for the two cameras. Secondary tasks were to check that this solution will also work in finding the actuators from the image and possibly allowing the bonded area measurements visually.
Evaluating the suitable illumination was performed by testing different sorts types of lighting illumination architectures and improving the most successful methods further. The first candidates were compared taking a set of sample images and calculating the fiber recognition performance for each.
Finally a new illumination module design was created designed to allow the use of special polarized backlight illumination inside the platform structure previously designed. This polarized illumination works using two crossed polarization filters behind and in front of the fibers to block direct light but letting through the light refracting and reflecting from the fibers thereby changing the type or direction of its polarization. The resulting image then has a good contrast between the black or dark background and the white or light coloured fiber targets.