Investigation and data adaptive analysis of CARS spectra

In this diploma work advantages of coherent anti-Stokes Raman scattering spectrometry (CARS) and various methods of the quantitative analysis of substance structure with its help are considered. The basic methods and concepts of the adaptive analysis are adduced. On the basis of these methods the algorithm of automatic measurement of a scattering strip size of a target component in CARS spectrum is developed. The algorithm uses known full spectrum of target substance and compares it with a CARS spectrum. The form of a differential spectrum is used as a feedback to control the accuracy of matching. To exclude the influence of a background in CARS spectra the differential spectrum is analysed by means of its second derivative. The algorithm is checked up on the simulated simple spectra and on the spectra of organic compounds received experimentally.

Unsupervised visual object categorization

The number of digital images has been increasing exponentially in the last few years. People have problems managing their image collections and finding a specific image. An automatic image categorization system could help them to manage images and find specific images. In this thesis, an unsupervised visual object categorization system was implemented to categorize a set of unknown images. The system is unsupervised, and hence, it does not need known images to train the system which needs to be manually obtained. Therefore, the number of possible categories and images can be huge. The system implemented in the thesis extracts local features from the images. These local features are used to build a codebook. The local features and the codebook are then used to generate a feature vector for an image. Images are categorized based on the feature vectors. The system is able to categorize any given set of images based on the visual appearance of the images. Images that have similar image regions are grouped together in the same category. Thus, for example, images which contain cars are assigned to the same cluster. The unsupervised visual object categorization system can be used in many situations, e.g., in an Internet search engine. The system can categorize images for a user, and the user can then easily find a specific type of image.

Soldiering and the making of Finnish manhood : conscription and masculinity in interwar Finland, 1918–1939

Military conscription and peacetime military service were the subjects of heated political, social and cultural controversies during the early years of national independence in Finland. Both the critics and the supporters of the existing military system described it as strongly formative of young men’s physical and moral development into adult men and male citizens. The conflicts over conscription prompted the contemporaries to express their notions about what Finnish men were like, at their best and at their worst, and what should and could be done about it. This thesis studies military conscription as an arena for the “making of manhood” in peacetime Finnish society, 1918–1939. It examines a range of public images of conscripted soldiering, asking how soldiering was depicted and given gendered meanings in parliamentary debates, war hero myths, texts concerned with the military and civic education of conscripts, as well as in works of fiction and reminiscences about military training as a personal experience. Studying conscription with a focus on masculinity, the thesis explores the different cultural images of manliness, soldiering and male citizenship on offer in Finnish society. It investigates how political parties, officers, educators, journalists, writers and “ordinary” conscripts used and developed, embraced or rejected these notions, according to their political purposes or personal needs. The period between the two world wars can be described as a fast-forward into military modernity in Finland. In the process, European middle class gender ideologies clashed with Finnish agrarian masculinities. Nationalistic agendas for the militarisation of Finnish manhood stumbled against intense class conflicts and ideological resistance. Military propaganda used images of military heroism, civic virtue and individual success to persuade the conscripts into ways of thinking and acting that were shaped by bourgeois mentality, nationalistic ideology and religious morality. These images are further analysed as expressive of the personal experiences and emotions of their middle-aged, male authors. The efforts of these military educators were, however, actively resisted on many fronts, ranging from rural working class masculinities among the conscripted young men to ideological critiques of the standing army system in parliament. In narratives about military training, masculinity was depicted as both strengthened and contradicted by the harsh and even brutal practices of interwar Finnish military training. The study represents a combination of new military history and the historical study of men and masculinities. It approaches masculinity as a contested and highly political form of social and cultural knowledge that is actively and selectively used by historic actors. Instead of trying to identify a dominant or “hegemonic” form of masculinity within a pre-determined theoretical structure, this study examines how the meanings ascribed to manhood varied according to class, age, political ideology and social situation. The interwar period in Finland can be understood as a period of contest between different notions of militarised masculinity, yet to judge by the materials studied, there was no clear winning party in that contest. A gradual movement from an atmosphere of conflict surrounding conscription towards political and cultural compromises can be discerned, yet this convergence was incomplete and many division lines remained.

Bag-of-Features Approach to Unsupervised Visual Object Categorisation

The large and growing number of digital images is making manual image search laborious. Only a fraction of the images contain metadata that can be used to search for a particular type of image. Thus, the main research question of this thesis is whether it is possible to learn visual object categories directly from images. Computers process images as long lists of pixels that do not have a clear connection to high-level semantics which could be used in the image search. There are various methods introduced in the literature to extract low-level image features and also approaches to connect these low-level features with high-level semantics. One of these approaches is called Bag-of-Features which is studied in the thesis. In the Bag-of-Features approach, the images are described using a visual codebook. The codebook is built from the descriptions of the image patches using clustering. The images are described by matching descriptions of image patches with the visual codebook and computing the number of matches for each code. In this thesis, unsupervised visual object categorisation using the Bag-of-Features approach is studied. The goal is to find groups of similar images, e.g., images that contain an object from the same category. The standard Bag-of-Features approach is improved by using spatial information and visual saliency. It was found that the performance of the visual object categorisation can be improved by using spatial information of local features to verify the matches. However, this process is computationally heavy, and thus, the number of images must be limited in the spatial matching, for example, by using the Bag-of-Features method as in this study. Different approaches for saliency detection are studied and a new method based on the Hessian-Affine local feature detector is proposed. The new method achieves comparable results with current state-of-the-art. The visual object categorisation performance was improved by using foreground segmentation based on saliency information, especially when the background could be considered as clutter.

AMB system for high-speed motors using automatic commissioning

The general trend towards increasing e ciency and energy density drives the industry to high-speed technologies. Active Magnetic Bearings (AMBs) are one of the technologies that allow contactless support of a rotating body. Theoretically, there are no limitations on the rotational speed. The absence of friction, low maintenance cost, micrometer precision, and programmable sti ness have made AMBs a viable choice for highdemanding applications. Along with the advances in power electronics, such as signi cantly improved reliability and cost, AMB systems have gained a wide adoption in the industry. The AMB system is a complex, open-loop unstable system with multiple inputs and outputs. For normal operation, such a system requires a feedback control. To meet the high demands for performance and robustness, model-based control techniques should be applied. These techniques require an accurate plant model description and uncertainty estimations. The advanced control methods require more e ort at the commissioning stage. In this work, a methodology is developed for an automatic commissioning of a subcritical, rigid gas blower machine. The commissioning process includes open-loop tuning of separate parts such as sensors and actuators. The next step is to apply a system identi cation procedure to obtain a model for the controller synthesis. Finally, a robust model-based controller is synthesized and experimentally evaluated in the full operating range of the system. The commissioning procedure is developed by applying only the system components available and a priori knowledge without any additional hardware. Thus, the work provides an intelligent system with a self-diagnostics feature and an automatic commissioning.

Synthesis and Characterization of Cellulose fiber-silica nanocomposites

Cellulose fiber-silica nanocomposites with novel mechanical, chemical and thermal properties have potential to be widely applied in different area. Monodispered silica nanoparticles play an important role in enhancing hybrids properties of hardness, strength, thermal stability etc. On the other hand, cellulose is one of the world’s most abundant and renewable polymers and possesses several unique properties required in many areas and biomedicine. The aim of this master thesis is to study if silica particles from reaction of sodium silicate and sulphuric acid can be adsorbed onto cellulose fiber surfaces via in situ growth. First, nanosilica particles were synthesized. Effect of pH and silica contents were tested. In theoretical part, introduction of silica, methods of preparation of nanosilica from sodium silicate, effect factors and additives were discussed. Then, cellulose fiber-silica nanocomposites were synthesis via route from sodium silicate and route silicic acid. In the experiment of route from sodium silicate, the effects of types of sodium silicate, pH and target ratio of silica to fiber were investigated. From another aspect, the effects of types of sodium silicate, fiber concentration in mixture solution and target ratio of silica to fiber were tested in the experiment of route from silicic acid. Samples were investigated via zeta potential measurement, particle size distribution, ash content measurement and Scanning Electron Microscopy (SEM). The Results of the experiment of preparing silica sol were that the particle size of silica sol was smaller prepared in pH 11.7 than that prepared in pH 9.3. Then in the experiment of synthesis of cellulose fiber-silica nanocomposites, it was concluded that the zeta potential of all the samples were around -16 mV and the highest ash content of all the samples was only 1.4%. The results of SEM images showed only a few of silica particles could be observed on the fiber surface, which corresponded to the value of ash content measurement.

Quantitative Magnetic Resonance Imaging of the Liver and Heart In Iron Overload

Systemic iron overload (IO) is considered a principal determinant in the clinical outcome of different forms of IO and in allogeneic hematopoietic stem cell transplantation (alloSCT). However, indirect markers for iron do not provide exact quantification of iron burden, and the evidence of iron-induced adverse effects in hematological diseases has not been established. Hepatic iron concentration (HIC) has been found to represent systemic IO, which can be quantified safely with magnetic resonance imaging (MRI), based on enhanced transverse relaxation. The iron measurement methods by MRI are evolving. The aims of this study were to implement and optimise the methodology of non-invasive iron measurement with MRI to assess the degree and the role of IO in the patients. An MRI-based HIC method (M-HIC) and a transverse relaxation rate (R2*) from M-HIC images were validated. Thereafter, a transverse relaxation rate (R2) from spin-echo imaging was calibrated for IO assessment. Two analysis methods, visual grading and rSI, for a rapid IO grading from in-phase and out-of-phase images were introduced. Additionally, clinical iron indicators were evaluated. The degree of hepatic and cardiac iron in our study patients and IO as a prognostic factor in patients undergoing alloSCT were explored. In vivo and in vitro validations indicated that M-HIC and R2* are both accurate in the quantification of liver iron. R2 was a reliable method for HIC quantification and covered a wider HIC range than M-HIC and R2*. The grading of IO was able to be performed rapidly with the visual grading and rSI methods. Transfusion load was more accurate than plasma ferritin in predicting transfusional IO. In patients with hematological disorders, the prevalence of hepatic IO was frequent, opposite to cardiac IO. Patients with myelodysplastic syndrome were found to be the most susceptible to IO. Pre-transplant IO predicted severe infections during the early post-transplant period, in contrast to the reduced risk of graft-versus-host disease. Iron-induced, poor transplantation results are most likely to be mediated by severe infections.