Master theses projects

1. Real-time rendering of accumulated snow
   Student: Per Ohlsson
   Supervisor: Stefan Seipel, Lars W. Pettersson
   Examiner: Stefan Seipel
   Publisher: Uppsala Master Theses in computer science 267.
   UU School of Engineering
   Abstract: This thesis presents a method of computing snow accumulation as a per pixel effect while rendering the scene. The method is similar to the shadow mapping method for shadow calculations. A depth buffer is used to find out how much snow a particular surface should receive. The amount of snow is then modified depending on the slope of the surface. To render the snow in a convincing way 3D noise is utilized for the lighting of the snow surface.

2. Computer-based morphometric assessment of spiral ganglion neurite outgrowth in vitro using image processing
   Student: Tomas Lundström and Henrik Boström
   Supervisor: Ingela Nyström
   Examiner: Ingela Nyström
   Partner: Dept. of Otolaryngology at UU Hospital
   Publisher: CBA Master Thesis No. 64, 75p., 2004,
   UU School of Engineering, UPTEC IT04 001
   Abstract: There is a pioneer research going on among medical researchers from all around the world; they want to make deaf people hear again in a natural way. To achieve this goal they want to make neuron-cells grow inside the human ear to reestablish the ability for deaf people to hear again.

   The research processes involves growing a huge amount of neurons in the laboratories and keeping track of the growth-rate and growth-behavior of these cells. There can be thousands of cells to keep track of every week. To perform these quantitative assessments of the growing cells the researchers started measuring the lengths of the neurite outgrowths, growing out from the seed of the cells.

   In this master thesis we have developed a digital image processing computer software for extracting and measuring neurite outgrowths in digital images.

   The processing of one such image involves four main digital image processing fields; these are thresholding, object classification, morphological operations and measuring by skeletonizing.

3. Leprechaun, a program for image analysis
   Student: Erik Andersson
   Supervisor: Carolina Wählby
   Examiner: Ewert Bengtsson
   Publisher: CBA Master Thesis No. 65, 35p., 2004,
   Abstract: This Master Thesis is about creating a program to handle image analysis in general with watershed segmentation as its main focus. The program is written in Java and is available for many platforms, such as Windows and Solaris. Many of the standard image analysis operations are implemented and extending the program is quite easy; menus and dialogs are created from xml-files and new commands can be inserted into a runnung program. The supported image types are grayscale, color, 3d and multi-spectral.

4. Multi-camera arrangement for automatic milking
   Student: Maria Pettersson and Johan Andrén Dinerf
   Supervisors: Fredrik Bergholm, Ingela Nyström, Anders Hallström
   Examiner: Fredrik Bergholm
   Partner: DeLaval International AB, Dept. VMS (Voluntary Milking System), Tumba
   Publisher: CBA Master Thesis No. 66, 94p., 2004,
   Abstract: The teat detection and positioning system used today on the DeLaval automatic milking system, VMS, comes with a number of drawbacks, that could be solved if it was replaced with a stereo vision system, placed outside the milking robot. This would decrease the damages on the present camera/laser detection device, and possibly increase the speed of the robot. This thesis is a feasibility study to find out if such a system is possible. The stereo calculations show that a stereo vision system is very sensitive. If such a system should work with high enough accuracy, the system needs to continuously be recalibrated, using reference points in the VMS. Results show that average error in absolute measurements is usually within the accepted range. The demand is higher when attaching a teat cup. Therefore relative measurements between objects in the picture is of higher interest. For example between a teat and the teat cup. Errors in relative measurements depend on the size of the relative measurement and is 8high. The image analysis does not detect the teats with high enough accuracy today, but shows that it is possible in an environment with appropriate illumination. All teats are seen using two stereo vision systems. Basics regarding both image analysis and robot milking are also presented. The main problem is divided into sub problems, which are investigated separately. To evaluate the methods, two extensive tests were performed. Final test 1 tests the actual stereo calculations while final test 2 tests the image analysis for teat detection. The final conclusion is that such a system is possible but is very sensitive. A final system needs to be more robust and exact. This project led to three Swedish patent applications.

5. Developement of the fiber orientation analyzer SPADES: a system using polarization-axis direction estimation
   Student: Simon Hensing
   Supervisor: Marco Lucisano, STFI-Packforsk AB, Stockholm
   Examiner: Gunilla Borgefors
   Partner: STFI-Packforsk AB, Stockholm
   Publisher: CBA Master Thesis No. 67, 31p., 2004,
   UU School of Engineering, UPTEC F04 022
   Abstract: The properties of a paper sheet is to a large extent dependent on the fiber orientation in the plane of the sheet. The purpose of this thesis is to create an on-line system for fiber orientation measuring based on the polarization effects of paper and to investigate wheter his technique can be implemented in the on-line system SOFA. The equipment consists of a polarization analyzer and uses a CCD-camera as light detector. Results show that the polarization axis of paper at visible wavelenghts correlates very well with the fiber orientation. The polarization effect is, however, quite limited and measurements require low noise levels. The conclusion is that the speed and accuracy of the system makes it a very competitive method for off-line fiber orientation analysis. However, the low noise levels required make it difficult to implement in SOFA and further development into an on-line system should be put on hold.

6. Automatic classification of images detected in Gyrolab
   Student: Pontus Olson
   Supervisor: Tobias Söderman, Gyros AB, Uppsala
   Examiner: Ewert Bengtsson
   Partner: Gyros AB, Uppsala
   Publisher: CBA Master Thesis No. 68, 37p., 2004,
   UU School of Engineering, UPTEC F04 044
   Abstract: Gyros AB is a biotechnical company which manufactures a system for protein quantification. Protein concentration is calculated from images produced from fluorescent molecules. Automatic classification of these images is desireable on a scale from poor to good, which indicates the quality of the preceding process if the image is suitable for protein quantification. In this thesis project, a classification system has been designed. Firstly, a set of parameters for the images has been constructed. Secondly, a neural network is used as a classifier. Results show that it is possible to a reasonable level of accuracy distinguish poor images from good images.

7. Image analysis as a tool for characterization of layering in stratified paper
   Student: Maria Sannes Lande
   Supervisor: Marco Lucisano, Ingela Nyström, Gunilla Borgefors
   Examiner: Gunilla Borgefors
   Partner: STFI-Packforsk AB, Stockholm
   Publisher: CBA Master Thesis No. 69, 66p., 2004,
   UU School of Engineering, UPTEC F04 057
   Abstract: The driving force in the paper industry is the ambition to make a paper that is both lighter and stronger than conventional paper of today. This may come true if paper has a layered structure, where the fibers in different layers have different properties. Producing such a paper is difficult at low basis weight. The forming method (i.e. the creation of the basic structure of the final paper) that works best, stratified forming has the disadvantage of layer mixing by which the fibers of the inner layer reach the surface of the paper. The study of the evolution of structure properties in the thickness direction of multiplayer paper is important to the design and optimization of machinery and processes for the commercial application of stratified forming. The goal of this project has been to develop tools to evaluate the quality of multi-layer paper based on image analysis to get information about the mixing of the layers. Three main questions were posed: 1; How do the layers mix? 2; How well do the outer layers cover the inner core? 3; How do flocs (fibers entangled in each other) move in the thickness direction? Although these questions have not fully been answered, I have developed methods that bring us a step closer to answering these questions. To study the paper on the inside, the paper has been split in thin layers. The paper has been produced in such a way that the fibers going into the inner and outer layers have been dyed differently instead of using fibers with different properties. To find out how the layers mix, a method was developed and programmed that identifies the fibers coming from the inner or outer layers and calculates the percentages of the two differently dyed fibers in the splits. To avoid user errors the program has been made so that the calculations are done automatically. To find out how well the outer layers cover the inner layer I have developed a program that rebuilds the paper digitally in the computer. The method works well. By studing flocs in the thin layers of the paper it is possible to see how these are spread in the paper and this might help to understand how flocs influence paper properties. A method that identifies flocs has been developed and the possibility to make volume images of flocs has been investigated.

8. Registration of tomographic animal volume images, from microPET, CT and MRT
   Student: Emma Gustafsson
   Supervisor: Mats Bergström, Uppsala Imanet AB
   Examiner: Ewert Bengtsson
   Partner: Uppsala Imanet AB
   Publisher: CBA Master Thesis No. 70, 26p., 2004,
   UU School of Engineering, UPTEC F04 063
   Abstract: Medical imaging is of great importance in many fields, both in clinical work and in medical research. Different imaging systems give different information about the patient, why it is valuable to combine/register different images to one another. The co-registrations allow precise comparisions of organs, anatomical regions and pathological processes between modalities. Many methods and programs have been developed for registration of human images, mostly brains, while little work has been done on full body animal images. Registration of animal images are of interest since many medical experiments are performed on rats or monkeys. This report describes the construction of a program performing registrations on animal images from three different modalities; PET, CT and MRT. The basis of the work have been another program, created for registration of human brain images. Changes and additions have been made to meet the requirements from this new field of application. Both global and local registrations have been used.

   Three experiments have been done to test the final program. The test images were from rats and a Marmoset monkey. The experiments showed that a method developed for human brain images can be used for full body registrations of animal images with a satisfying result, especially if the images are from the same animal. When the images are from different individuals the results are a little poorer, but still fairly good.

9. Layer segmentation in cross-section images of board
   Student: Ingemar Holmqvist
   Supervisor: Gunilla Borgefors
   Examiner: Gunilla Borgefors
   Partner: Örjan Sävborg, StoraEnso Research, Falun
   Publisher: CBA Master Thesis No. 71, 53p., 2004,
   UU School of Engineering, UPTEC F04 067
   Abstract: Paper is one of the most common substances in the world having literaly thousand of uses - from tissues to cardboard boxes. In the 3D Tracking of fibres in Paper-project (1997-2002, within the National VISIT programme) a small digital volume of three layer paper board, was meticulously created through the use of an electron microscope with the idea that image analysis could be used to further our knowledge about the inner structure of paper. This volume consists of 102 consecutive high resolution cross-section images showing the fibre structure of the paper. In this thesis, a layer segmentation algorithm based on simple image processing techniques such as filtering, edge detection and morphological operations is developed with the intent of separating the volume into its three different layers. We beging by filtering the images to remove the fibre structure and access the hidden inner layer structure. Edge detection algorithms are then used on the filtered images to extract possible layer borders. Depending on the a prirori knowledge of the layer borders different techniques based both on morphological operations and interpolation are used to extract the best possible border candidates. Experiments conducted both on the whole volume set and separate reference images shows that developed methods are both powerful and accurate.

10. Automatic method for acquiring paper cross section images using a scanning electron microscope
    Student: Åsa Odell
    Supervisor: Örjan Sävborg, StoraEnso Research, Falun
    Examiner: Gunilla Borgefors
    Partner: StoraEnso Research, Falun
    Publisher: CBA Master Thesis No. 72, 30p., 2004,
    UU School of Engineering, UPTEC F04 076
    Abstract: Stora Enso Research center Falun performs research on paper. Small cross sections of paper, about two cm long, are sequentially viewed and photographed using a scanning electron microscope, SEM. The images from the SEM are stored on the hard drive of the microscope computer and are then analyzed. Each paper sample produces about 100 to 150 images. For each acquiring of an image, the operator of the microscope has to perform certain image settings and operations through clicks and scrolling with the computer mouse. A series of about 100 sample images demands large amounts of time when many operations must be performed per image. In fact it takes several hours to acquire all images from just one sample. In order to decrease the amount of work for Stora Enso staff, through lessening the need for staff to be present at the SEM during image acquisition and also to speed up the analyzing process to win time, thereby increasing cost efficiency, a method to automate the process of image acquiring has been created. To accomplish this task image analysis and computer communication were used as the main tools. Image analysis acts as a virtual eye to determine characteristic and/or critical points in a SEM image for decision making. Computer communication is used for commanding the SEM to perform certain actions. Combining these tools, a program acquiring images without human intervention was created and hidden behind a user friendly interface. The program was tested on many different kinds of paper. It could be concluded that the total time demanded for the acquisition of a series of images, was drastically reduced. A series of a total of 100 images of any sample type can now take a little less than 1 h to acquire, as opposed to the several hours before, from the moment when the first scanning of an image starts, and no staff needs to be present at the SEM during that time.

11. Real-time surface rendering for interactive volume image segmentation in a haptic environment
    Student: Jonas Agmund
    Supervisor: Erik Vidholm and Ewert Bengtsson
    Examiner: Ewert Bengtsson
    Publisher: CBA Master Thesis No. 73, 35p., 2004,
    UU School of Engineering, UPTEC F04 071
    Abstract: Volume image segmentation is a very important step when analyzing medical volume images. This Masters thesis describes the implementation of a fast surface rendering algorithm that allows interactive volume image segmentation in a haptic environment. The implementation uses a highly optimized marching cubes algorithm which is made efficient by dividing it into two major parts, surface extraction and triangle generation. The surface extraction is implemented by an efficient surface tracking algorithm that avoids searching empty space. The resulting surface information is used in conjunction with an intelligent caching strategy for fast triangle generation. The implementation of the surface renderer in the haptic environment makes it possible to achieve real-time frame rates while editing segmented objects guided by haptic feedback. Segmentation by thresholding has been implemented along with basic editing tools such as drawing, erasing, erosion and dilation. The surface renderer has shown to be efficient for arbitrary volume image sizes, and allows interactive segmentation and manipulation of moderately sized volumes.

12. Visualization of the three dimensional fibre stucture of paper
    Student: Erik Cedheim
    Supervisor: Gunilla Borgefors
    Examiner: Gunilla Borgefors
    Partner: Örjan Sävborg, StoraEnso Research, Falun
    Publisher: CBA Master Thesis No. 74, 56p., 2004,
    UU Master Thesis in Computing Science
    Abstract: This master thesis focuses on visualization of the three dimensional fibre stucture of paper-board. The data was originally captured by StoraEnso using Scanning Electron Microscopy (SEM) and was later assembled by Mattias Aronsson into a 3D structure. The assembled volume has a low resolution in the z-direction making visualization as well as analysis a difficult task. Included in the task is the visualization of three different layers, originally segmented by Ingemar Holmkvist. The program is developed so that the user can interactively rotate and manipulate the volum in real-time. The program structure is focused on speed and memory efficiency rather than perfect image quality which makes the final product highly interactive with a good visualization of the fibre stucture.