The genetical and environmental control of a number of characters of agronomic importance in barley has been investigated by triple test cross (TTC) and linear model-fitting analyses. Additive and dominance genetic effects are observed for height, maturity, yield and yield component characters with the exception of tiller number, which was found to be almost totally controlled by environmental factors. Epistasis was found to be important only for grain number, although it was observed in other characters (notably neck length). Other analyses on the same data set have also been made to detect the presence of genotype x environment interactions and linkage in the above characteristics, together with phenotypic and genotypic correlations between them. Only height at harvest and 1000 grain weight appeared to have potential for early generation selection. As a result, efficient prediction of the potential of a particular cross for the majority of characters of agronomic importance is shown to be necessary and important. The second part of this work involved testing the effectiveness of cross-prediction methods based on the results of the genetical analyses. It is shown that the potential of a cross to produce superior inbred lines can be successfully predicted from TTC and model-fitting analyses. The prediction methods have been shown to be successful in identifying the cross from a number of crosses with the greatest potential in respect of both single characters and pairs of characters. It is shown that yield itself can be studied successfully in this way, as well as those characters with high heritability such as height and 1000 grain weight. Furthermore it is shown that an estimate of the additive genetic variance necessary for making the predictions, of similar accuracy to that obtained from the TTC, can be Je<-ive<J_ from an F family analysis, thus saving considerably on the experimental work necessary. The predictions are compared with the results of a normal selection programme practised on the same material to illustrate the need for early generation prediction and delay of selection until later generations to prevent loss of potentially useful lines in early generations.

No description available.

No description available.

<p>A proposal for a person authentication system, which localizes facial landmarks and extracts biometrical features for face authentication, is presented in this thesis. An efficient algorithm for eye localization and biometrical feature extraction and person identification is developed by using Gabor filters. In the eye localization part, we build artificial average eye models for eye location. In the person identification part, we construct databases of biometrical features around the eye area of clients and, for authentication, Schwartz inequality and the sum square error (SSE) are used.</p><p>This project is implemented in the ‘Matlab’ programming language, on a personal computer system, and experimental results on the proposed system are presented.</p>

A proposal for a person authentication system, which localizes facial landmarks and extracts biometrical features for face authentication, is presented in this thesis. An efficient algorithm for eye localization and biometrical feature extraction and person identification is developed by using Gabor filters. In the eye localization part, we build artificial average eye models for eye location. In the person identification part, we construct databases of biometrical features around the eye area of clients and, for authentication, Schwartz inequality and the sum square error (SSE) are used. This project is implemented in the ‘Matlab’ programming language, on a personal computer system, and experimental results on the proposed system are presented.

The research investigates fingerprint recognition as one of the most reliable biometrics identification methods. An automatic identification process of humans-based on fingerprints requires the input fingerprint to be matched with a large number of fingerprints in a database. To reduce the search time and computational complexity, it is desirable to classify the database of fingerprints into an accurate and consistent manner so that the input fingerprint is matched only with a subset of the fingerprints in the database. In this regard, the research addressed fingerprint classification. The goal is to improve the accuracy and speed up of existing automatic fingerprint identification algorithms. The investigation is based on analysis of fingerprint characteristics and feature classification using neural network and fuzzy-neural classifiers. The methodology developed, is comprised of image processing, computation of a directional field image, singular-point detection, and feature vector encoding. The statistical distribution of feature vectors was analysed using SPSS. Three types of classifiers, namely, multi-layered perceptrons, radial basis function and fuzzy-neural methods were implemented. The developed classification systems were tested and evaluated on 4,000 fingerprint images on the NIST-4 database. For the five-class problem, classification accuracy of 96.2% for FNN, 96.07% for MLP and 84.54% for RBF was achieved, without any rejection. FNN and MLP classification results are significant in comparison with existing studies, which have been reviewed.

Orientador: José Mario De Martino / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-23T20:06:41Z (GMT). No. of bitstreams: 1 Pereira_TiagodeFreitas_M.pdf: 17638731 bytes, checksum: 15a8d07214e3b31accd3218e5bde20cb (MD5) Previous issue date: 2013 / Resumo: O Resumo poderá ser visualizado no texto completo da tese digital / Abstract: The complete Abstract is available with the full electronic document. / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

No description available.

M.Sc. (Computer Science) / An increasing number of corporate and government institutions are utilising electronic commerce to provide or improve their services. These new online services are becoming increasingly complex, offering diverse functionality and managing high volumes of personal and confidential data. The protection and confidentiality of such data is imperative but the security mechanisms and the policies governing its security are rarely sufficient. Nonetheless electronic commerce service providers market their services as being “secure” and by doing so they are developing a false sense of security within computer users. Average computer users are aware of security threats like hackers, viruses, Trojans and spyware, but their limited computer knowledge doesn’t allow them to understand, identify or respond to such security threats. A lack of computer knowledge, little experience and gullibility render the average computer user incapable of managing computer security. This is even more true when the average computer user is put up against the wit and cunning of a hacker. Electronic commerce has changed, the threats have changed, the users have changed and electronic commerce security solutions remain the same. Hackers are no longer hobbyists: they hack for financial gain and not fame, they work together and they exploit any security weakness to get what they want. More and more often the average computer user falls victim to hacker attacks, not only because of the above mentioned human factors but also because of weak security mechanisms that govern users’ access to critical online services.

Biomimetics has traditionally meant the study of the structure and function of biological systems as models for the design and engineering of materials and machines. Some stunning results showing how biomimetics has been put into practice include Velcro, the Millennium Dome in London and the Millennium Bridge in spanning the River Tyne in Newcastle. Business biomimetics is a new term developed as part of this thesis and is proposed as an emerging field within biomimetics. It is distinct from the current use of biomimetics in that it demonstrates a strategic use in the business environment. This thesis examines how biomimetics can be used to aid business and tests whether businesses can improve their fitness for the future by applying the principles of biomimetics to the development of business strategy. Research groups at the University of Bath were used as the basis of testing this hypothesis, fitness for the future. Results were measured by comparing conventional management models with newly developed business biomimetic models. Analysis of the results demonstrated evidence to support the hypothesis that business biomimetics had a positive effect on the future fitness for the groups as measured by increases in values for esteem and the generation of intellectual capital. These are nationally recognised measures for the success of research groups. Further evidence was gathered from a number of case studies that span a range of industries and disciplines. These case studies show that results are encouraging and the ideas generated by the business biomimetic models demonstrate a richer set of ideas that inspire the managers involved. One case follows the business from previous failure through the development of business biomimetically inspired ideas, implementation and into measureable results. Subsequent improvements include increased revenue, reduced costs, improved supply chain efficiency, better customer relationships and a clearer understanding of the cross functional integration value. The delivery of a business biomimetic approach was facilitated by my invention of the Infinity Context Free Process that provided a framework to translate business problems into biological research areas and then the formulation of implementable business projects. A glimpse into the future for business biomimetics is provided in Chapter 9 that presents examples of the application of business biomimetics and shares the embryonic first steps by businesses to use nature as a guiding principle in sustainability, corporate social responsibility, growth and leadership development. The use of business biomimetics opens up the 3.8 billion years worth of nature‘s expertise and provides a new way for business leaders to innovate. The thesis concludes that when coupled with conventional management approaches, it further provides a way to design and manage businesses that are fit for now and for the future.