Automatic traffic sign inventory- and condition analysis

This thesis researches automatic traffic sign inventory and condition analysis using machine vision and pattern recognition methods. Automatic traffic sign inventory and condition analysis can be used to more efficient road maintenance, improving the maintenance processes, and to enable intelligent driving systems. Automatic traffic sign detection and classification has been researched before from the viewpoint of self-driving vehicles, driver assistance systems, and the use of signs in mapping services. Machine vision based inventory of traffic signs consists of detection, classification, localization, and condition analysis of traffic signs. The produced machine vision system performance is estimated with three datasets, from which two of have been been collected for this thesis. Based on the experiments almost all traffic signs can be detected, classified, and located and their condition analysed. In future, the inventory system performance has to be verified in challenging conditions and the system has to be pilot tested.

Dual light barrier method for six degrees of freedom tool center point calibration of an industrial robot

Tool center point calibration is a known problem in industrial robotics. The major focus of academic research is to enhance the accuracy and repeatability of next generation robots. However, operators of currently available robots are working within the limits of the robot´s repeatability and require calibration methods suitable for these basic applications. This study was conducted in association with Stresstech Oy, which provides solutions for manufacturing quality control. Their sensor, based on the Barkhausen noise effect, requires accurate positioning. The accuracy requirement admits a tool center point calibration problem if measurements are executed with an industrial robot. Multiple possibilities are available in the market for automatic tool center point calibration. Manufacturers provide customized calibrators to most robot types and tools. With the handmade sensors and multiple robot types that Stresstech uses, this would require great deal of labor. This thesis introduces a calibration method that is suitable for all robots which have two digital input ports free. It functions with the traditional method of using a light barrier to detect the tool in the robot coordinate system. However, this method utilizes two parallel light barriers to simultaneously measure and detect the center axis of the tool. Rotations about two axes are defined with the center axis. The last rotation about the Z-axis is calculated for tools that have different width of X- and Y-axes. The results indicate that this method is suitable for calibrating the geometric tool center point of a Barkhausen noise sensor. In the repeatability tests, a standard deviation inside robot repeatability was acquired. The Barkhausen noise signal was also evaluated after recalibration and the results indicate correct calibration. However, future studies should be conducted using a more accurate manipulator, since the method employs the robot itself as a measuring device.

Systemaattinen layout-suunnittelu

Layout-suunnittelu on tyypillisesti aikaa vievä prosessi, jossa joudutaan huomioimaan useita erilaisia muuttujia. Tässä työssä esitellään layout-suunnittelun helpottamiseksi kehitettyjä systemaattisia menetelmiä, ja arvioidaan niiden käyttömahdollisuuksia.

Combined selection in early generation testing of self-pollinated plants

Seven selection indexes based on the phenotypic value of the individual and the mean performance of its family were assessed for their application in breeding of self-pollinated plants. There is no clear superiority from one index to another although some show one or more negative aspects, such as favoring the selection of a top performing plant from an inferior family in detriment of an excellent plant from a superior family

The interaction of meal-related, rhythmic and homeostatic mechanisms and the generation of thirst and drinking

One of the primary goals of the study of thirst is to understand why drinking occurs under ad libitum or natural conditions. An appreciation of the experimental strategies applied by physiologists studying thirst from different perspectives can facilitate progress toward understanding the natural history of drinking behavior. Drinking research carried out using three separate perspectives - homeostatic, circadian rhythms, and food-associated - generates types of information about the mechanisms underlying drinking behavior. By combining research strategies and methods derived from each of these approaches, it has been possible to gain new information that increases our appreciation of the interactions between homeostatic mechanisms and circadian rhythms in the modulation of water intake and how these might be related to drinking associated with food intake under near natural conditions

Low-cost automatic activity data recording system

We describe a low-cost, high quality device capable of monitoring indirect activity by detecting touch-release events on a conducting surface, i.e., the animal's cage cover. In addition to the detecting sensor itself, the system includes an IBM PC interface for prompt data storage. The hardware/software design, while serving for other purposes, is used to record the circadian activity rhythm pattern of rats with time in an automated computerized fashion using minimal cost computer equipment (IBM PC XT). Once the sensor detects a touch-release action of the rat in the upper portion of the cage, the interface sends a command to the PC which records the time (hours-minutes-seconds) when the activity occurred. As a result, the computer builds up several files (one per detector/sensor) containing a time list of all recorded events. Data can be visualized in terms of actograms, indicating the number of detections per hour, and analyzed by mathematical tools such as Fast Fourier Transform (FFT) or cosinor. In order to demonstrate method validation, an experiment was conducted on 8 Wistar rats under 12/12-h light/dark cycle conditions (lights on at 7:00 a.m.). Results show a biological validation of the method since it detected the presence of circadian activity rhythm patterns in the behavior of the rats

The rate of force generation by the myocardium is not influenced by afterload

The influence of afterload on the rate of force generation by the myocardium was investigated using two types of preparations: the in situ dog heart (dP/dt) and isolated papillary muscle of rats (dT/dt). Thirteen anesthetized, mechanically ventilated and thoracotomized dogs were submitted to pharmacological autonomic blockade (3.0 mg/kg oxprenolol plus 0.5 mg/kg atropine). A reservoir connected to the left atrium permitted the control of left ventricular end-diastolic pressure (LVEDP). A mechanical constriction of the descending thoracic aorta allowed to increase the systolic pressure in two steps of 20 mmHg (conditions H1 and H2) above control values (condition C). After arterial pressure elevations (systolic pressure C: 119 ± 8.1; H1: 142 ± 7.9; H2 166 ± 7.7 mmHg; P<0.01), there were no significant differences in heart rate (C: 125 ± 13.9; H1: 125 ± 13.5; H2: 123 ± 14.1 bpm; P>0.05) or LVEDP (C: 6.2 ± 2.48; H1: 6.3 ± 2.43; H2: 6.1 ± 2.51 mmHg; P>0.05). The values of dP/dt did not change after each elevation of arterial pressure (C: 3,068 ± 1,057; H1: 3,112 ± 996; H2: 3,086 ± 980 mmHg/s; P>0.05). In isolated rat papillary muscle, an afterload corresponding to 50% and 75% of the maximal developed tension did not alter the values of the maximum rate of tension development (100%: 78 ± 13; 75%: 80 ± 13; 50%: 79 ± 11 g mm-2 s-1, P>0.05). The results show that the rise in afterload per se does not cause changes in dP/dt or dT/dt

Algorithmic Techniques in Gene Expression Processing. From Imputation to Visualization

The amount of biological data has grown exponentially in recent decades. Modern biotechnologies, such as microarrays and next-generation sequencing, are capable to produce massive amounts of biomedical data in a single experiment. As the amount of the data is rapidly growing there is an urgent need for reliable computational methods for analyzing and visualizing it. This thesis addresses this need by studying how to efficiently and reliably analyze and visualize high-dimensional data, especially that obtained from gene expression microarray experiments. First, we will study the ways to improve the quality of microarray data by replacing (imputing) the missing data entries with the estimated values for these entries. Missing value imputation is a method which is commonly used to make the original incomplete data complete, thus making it easier to be analyzed with statistical and computational methods. Our novel approach was to use curated external biological information as a guide for the missing value imputation. Secondly, we studied the effect of missing value imputation on the downstream data analysis methods like clustering. We compared multiple recent imputation algorithms against 8 publicly available microarray data sets. It was observed that the missing value imputation indeed is a rational way to improve the quality of biological data. The research revealed differences between the clustering results obtained with different imputation methods. On most data sets, the simple and fast k-NN imputation was good enough, but there were also needs for more advanced imputation methods, such as Bayesian Principal Component Algorithm (BPCA). Finally, we studied the visualization of biological network data. Biological interaction networks are examples of the outcome of multiple biological experiments such as using the gene microarray techniques. Such networks are typically very large and highly connected, thus there is a need for fast algorithms for producing visually pleasant layouts. A computationally efficient way to produce layouts of large biological interaction networks was developed. The algorithm uses multilevel optimization within the regular force directed graph layout algorithm.

Precipitated immune complexes of IgM induce the generation of reactive oxygen species by rabbit polymorphonuclear leucocytes

We report that immune complexes of IgM (ICIgM) antibodies and ovalbumin in the form of a precipitate from the equivalence zone induce the generation of reactive oxygen species by rabbit blood polymorphonuclear leucocytes (PMN), as measured by the chemiluminescence (CL) production in the presence of luminol. The kinetics of CL generation induced by ICIgM is quite different from that induced by precipitated immune complexes of IgG (ICIgG): the maximum rate of CL production for ICIgM occurs around 14 min, whereas for ICIgG it occurs about 5 min after incubation with the cells. Also the triggering of the process requires a higher concentration of ICIgM than of ICIgG. Evidence is presented that these effects are not mediated by interaction of the antigen (ovalbumin) with the cell, since immune precipitates of ovalbumin and the F(ab')2 fragment had no effect. Our observations that precipitated ICIgM can also be an effective stimulus for CL generation and thus for O2- production reveal a new functional capability of PMN. These results may have implications for the understanding of the participation of ICIgM (as well as of ICIgG) in inflammatory reactions mediated by PMN in immune complex diseases, and in the mechanisms of defense against microbes and other non-self agents.

Establishment of new murine embryonic stem cell lines for the generation of mouse models of human genetic diseases

Embryonic stem cells are totipotent cells derived from the inner cell mass of blastocysts. Recently, the development of appropriate culture conditions for the differentiation of these cells into specific cell types has permitted their use as potential therapeutic agents for several diseases. In addition, manipulation of their genome in vitro allows the creation of animal models of human genetic diseases and for the study of gene function in vivo. We report the establishment of new lines of murine embryonic stem cells from preimplantation stage embryos of 129/Sv mice. Most of these cells had a normal karyotype and an XY sex chromosome composition. The pluripotent properties of the cell lines obtained were analyzed on the basis of their alkaline phosphatase activity and their capacity to form complex embryoid bodies with rhythmically contracting cardiomyocytes. Two lines, USP-1 and USP-3, with the best in vitro characteristics of pluripotency were used in chimera-generating experiments. The capacity to contribute to the germ line was demonstrated by the USP-1 cell line. This cell line is currently being used to generate mouse models of human diseases.

A third generation regimen VACOP-B with or without adjuvant radiotherapy for aggressive localized non-Hodgkin's lymphoma: report from the Italian Non-Hodgkin's Lymphoma Co-operative Study Group

The objective of this multicenter prospective study was to determine the clinical efficacy and toxicity of a polychemotherapeutic third generation regimen, VACOP-B, with or without radiotherapy as front-line therapy in aggressive localized non-Hodgkin's lymphoma. Ninety-three adult patients (47 males and 46 females, median age 45 years) with aggressive localized non-Hodgkin's lymphoma, 43 in stage I and 50 in stage II (non-bulky), were included in the study. Stage I patients received VACOP-B for 6 weeks plus involved field radiotherapy and stage II patients received 12 weeks VACOP-B plus involved field radiotherapy on residual masses. Eighty-six (92.5%) achieved complete remission and 4 (4.3%) partial remission. Three patients (3.2%) were primarily resistant. Ten-year probability of survival, progression-free survival and disease-free survival were 87.3, 79.9 and 83.9%, respectively. Eighty-four patients are surviving at a median observation time of 57 months (range: 6-126). Statistical analysis showed no difference between stages I and II in terms of response, ten-year probability of survival, progression-free survival or disease-free survival. Side effects and toxicity were negligible and were similar in the two patient groups. The results of this prospective study suggest that 6 weeks of VACOP-B treatment plus radiotherapy may be the therapy of choice in stage I aggressive non-Hodgkin's lymphoma. Twelve weeks of VACOP-B treatment with or without radiotherapy was shown to be effective and feasible for stage II. These observations need to be confirmed by a phase III study comparing first and third generation protocols in stage I-II aggressive non-Hodgkin's lymphoma.

A semi-automatic computerized method to measure baroreflex-mediated heart rate responses that reduces interobserver variability

Arterial baroreflex sensitivity estimated by pharmacological impulse stimuli depends on intrinsic signal variability and usually a subjective choice of blood pressure (BP) and heart rate (HR) values. We propose a semi-automatic method to estimate cardiovascular reflex sensitivity to bolus infusions of phenylephrine and nitroprusside. Beat-to-beat BP and HR time series for male Wistar rats (N = 13) were obtained from the digitized signal (sample frequency = 2 kHz) and analyzed by the proposed method (PRM) developed in Matlab language. In the PRM, time series were low-pass filtered with zero-phase distortion (3rd order Butterworth used in the forward and reverse direction) and presented graphically, and parameters were selected interactively. Differences between basal mean values and peak BP (deltaBP) and HR (deltaHR) values after drug infusions were used to calculate baroreflex sensitivity indexes, defined as the deltaHR/deltaBP ratio. The PRM was compared to the method traditionally (TDM) employed by seven independent observers using files for reflex bradycardia (N = 43) and tachycardia (N = 61). Agreement was assessed by Bland and Altman plots. Dispersion among users, measured as the standard deviation, was higher for TDM for reflex bradycardia (0.60 ± 0.46 vs 0.21 ± 0.26 bpm/mmHg for PRM, P < 0.001) and tachycardia (0.83 ± 0.62 vs 0.28 ± 0.28 bpm/mmHg for PRM, P < 0.001). The advantage of the present method is related to its objectivity, since the routine automatically calculates the desired parameters according to previous software instructions. This is an objective, robust and easy-to-use tool for cardiovascular reflex studies.

Automatic image-based identification of Saimaa ringed seals

The Saimaa ringed seal is one of the most endangered seals in the world. It is a symbol of Lake Saimaa and a lot of effort have been applied to save it. Traditional methods of seal monitoring include capturing the animals and installing sensors on their bodies. These invasive methods for identifying can be painful and affect the behavior of the animals. Automatic identification of seals using computer vision provides a more humane method for the monitoring. This Master's thesis focuses on automatic image-based identification of the Saimaa ringed seals. This consists of detection and segmentation of a seal in an image, analysis of its ring patterns, and identification of the detected seal based on the features of the ring patterns. The proposed algorithm is evaluated with a dataset of 131 individual seals. Based on the experiments with 363 images, 81\% of the images were successfully segmented automatically. Furthermore, a new approach for interactive identification of Saimaa ringed seals is proposed. The results of this research are a starting point for future research in the topic of seal photo-identification.

Automatic recognition of fish from video sequences

The problem of automatic recognition of the fish from the video sequences is discussed in this Master’s Thesis. This is a very urgent issue for many organizations engaged in fish farming in Finland and Russia because the process of automation control and counting of individual species is turning point in the industry. The difficulties and the specific features of the problem have been identified in order to find a solution and propose some recommendations for the components of the automated fish recognition system. Methods such as background subtraction, Kalman filtering and Viola-Jones method were implemented during this work for detection, tracking and estimation of fish parameters. Both the results of the experiments and the choice of the appropriate methods strongly depend on the quality and the type of a video which is used as an input data. Practical experiments have demonstrated that not all methods can produce good results for real data, whereas on synthetic data they operate satisfactorily.