Thesis (MEng)--Stellenbosch University, 2000. / ENGLISH ABSTRACT: The purpose of this thesis is to investigate the nature of digital image compression and the calculation of the quality of the compressed images. The work is focused on greyscale images in the domain of satellite images and aerial photographs. Two compression techniques are studied in detail namely the JPEG and fractal compression methods. Implementations of both these techniques are then applied to a set of test images. The rest of this thesis is dedicated to investigating the measurement of the loss of quality that was introduced by the compression. A general method for quality measurement (signal To Noise Ratio) is discussed as well as a technique that was presented in literature quite recently (Grey Block Distance). Hereafter, a new measure is presented. After this, a means of comparing the performance of these measures is presented. It was found that the new measure for image quality estimation performed marginally better than the SNR algorithm. Lastly, some possible improvements on this technique are mentioned and the validity of the method used for comparing the quality measures is discussed. / AFRIKAANSE OPSOMMING: Die doel van hierdie tesis is om ondersoek in te stel na die aard van digitale beeldsamepersing en die berekening van beeldkwaliteit na samepersing. Daar word gekonsentreer op grysvlak beelde in die spesifieke domein van satellietbeelde en lugfotos. Twee spesifieke samepersingstegnieke word in diepte ondersoek naamlik die JPEG en fraktale samepersingsmetodes. Implementasies van beide hierdie tegnieke word op 'n stel toetsbeelde aangewend. Die res van hierdie tesis word dan gewy aan die ondersoek van die meting van die kwaliteitsverlies van hierdie saamgeperste beelde. Daar word gekyk na 'n metode wat in algemene gebruik in die praktyk is asook na 'n nuwer metode wat onlangs in die literatuur veskyn het. Hierna word 'n nuwe tegniek bekendgestel. Verder word daar 'n vergelyking van hierdie mates en 'n ondersoek na die interpretasie van die 'kwaliteit' van hierdie kwaliteitsmate gedoen. Daar is gevind dat die nuwe maatstaf vir kwaliteit net so goed en selfs beter werk as die algemene maat vir beeldkwaliteit naamlik die Sein tot Ruis Verhouding. Laastens word daar moontlike verbeterings op die maatstaf genoem en daar volg 'n bespreking oor die geldigheid van die metode wat gevolg is om die kwaliteit van die kwaliteitsmate te bepaal

Thesis (MScEng)--University of Stellenbosch, 2001. / ENGLISH ABSTRACT: One of the problems with any deep space communication system is that information may be altered or lost during transmission due to channel noise. It is known that any damage to the bit stream may lead to objectionable visual quality distortion of images at the decoder. The purpose of this thesis is to design an error correction and data compression algorithm for image protection, which will allow the communication bandwidth to be better utilized. The work focuses on Sunsat (Stellenbosch Satellite) images as test images. Investigations were done on the JPEG 2000 compression algorithm's robustness to random errors, putting more emphasis on how much of the image is degraded after compression. Implementation of both the error control coding and data compression strategy is then applied to a set of test images. The FEe algorithm combats some if not all of the simulated random errors introduced by the channel. The results illustrates that the error correction of random errors is achieved by a factor of 100 times (xl00) on all test images and that the probability of error of 10-2in the channel (10-4for image data) shows that the errors causes little degradation on the image quality. / AFRIKAANSE OPSOMMING: Een van die probleme met kommunikasie in die ruimte is dat informasie mag verlore gaan en! of gekorrupteer word deur ruis gedurende versending deur die kanaal. Dit is bekend dat enige skade aan die bisstroom mag lei tot hinderlike vervorming van die beelde wat op aarde ontvang word. Die doel van hierdie tesis om foutkorreksie en datakompressie te ontwikkel wat die satelliet beelde sal beskerm gedurende versending en die kommunikasie kanaal se bandwydte beter sal benut. Die werk fokus op SUNSAT (Stellenbosch Universiteit Satelliet) se beelde as toetsbeelde. Ondersoeke is gedoen na die JPEG2000 kompressie algoritme se bestandheid teen toevalsfoute, met klem op hoeveel die beeld gedegradeer word deur die bisfoute wat voorkom. Beide die kompressie en die foutkorreksie is ge-implementeer en aangewend op die toetsbeelde. Die foutkorreksie bestry die gesimuleerde toevalsfoute, soos wat dit op die kanaal voorkom. Die resultate toon dat die foutkorreksie die toevalsfoute met 'n faktor 100 verminder, en dat 'n foutwaarskynlikheid van 10-2 op die kanaal (10-4 op die beelddata) weinig degradering in die beeldkwaliteit veroorsaak.

Thesis (MScEng)--University of Stellenbosch, 2002. / ENGLISH ABSTRACT: The primary payload of the University of Stellenbosch's micro-satellite, SUNSAT I, was a high quality imager, capable of taking stereoscopic images of the surface of the earth. Although the orbit of SUNSAT I will have the satellite pass over the whole earth, contact via the SUNSAT I ground station at the University will only be made for 3% of each day. To be able to photograph any part of the earth onboard memory was provided on the satellite to store the image until it can be downloaded to the ground station. A high speed downlink was also added to the satellite to be able to download a complete image from the onboard memory in one pass and also to take realtime pictures as the satellite passes over the ground station. At that stage there was no way to capture the data at the ground station. A high speed digital data capture mechanism was needed. The thesis discusses the development of a high speed capture card. Due to the high speeds needed it provided an excellent opportunity to develop the card using the PCI bus, the first design to do so at the University, as an interface with a computer. A prototype card was developed first to do proof of concept. It is shown that the prototype card fulfilled the functional requirements and it was also used to capture the first image from the complete satellite during the final tests before launch. Finally a production card for use in the ground station was designed and assembled. / AFRIKAANSE OPSOMMING: Die hoof loonvrag van die Universtieit van Stellenbosch se eerste mikrosatelliet, SUNSAT I, is "n hoë kwaliteit kamera wat stereoskopiese fotos van die oppervlak van die aarde kan neem. Alhoewel die satelliet oor die hele oppervlakte van die aarde gaan beweeg, is dit vir slegs 3% van die dag sigbaar vanaf die grondstasie by die Universiteit van Stellenbosch. Om 'n foto van enige plek op aarde te kan neem is daar aanboord geheue op die satelliet geplaas om die foto te stoor totdat dit by die grondsatsie afgelaai kan word. "n Hoë spoed skakel is ook daargestelom die data in die geheue van die satelliet in een verbyvlug te kan aflaai. Dit stel die satelliet ook in staat om intydse fotos gedurende In verbyvlug van die satelliet te kan neem. Op daardie stadium het daar nog nie "n manier bestaan om die hoë spoed data vas te lê nie. In Meganisme om die hoë spoed digitale data te vang was nodig. Hierdie tesis bespreek die ontwikkeling van so In data vang kaart. Weens die hoë snelhede wat benodig word, het die PCI bus die ideale oplossing gebied om die data teen "n hoë spoed op In persoonlike rekenaar te stoor. Dit was ook die eerste ontwerp aan die Universiteit wat van die PCI bus gebruik gemaak het. In die eerste instansie is In prototipe ontwikkel om te bewys dat dit wel moontlik is om die data teen die benodigde tempo te kan stoor. Daar word gewys dat die prototipe aan die behoefte voldoen en dit is ook gebruik om die eerste beeld vanaf die volledige satelliet te vang voordat dit gelanseer is. In die tweede instansie word In produksiekaart ontwikkel en aanmekaar gesit.

Thesis (MScEng)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: South Africa's first micro satellite, SUNSAT, was operational in orbit for 23 months since its NASA-sponsored launch on February 23, 1999. SUNSAT is a graduate studentdeveloped satellite from the University of Stellenbosch in South Africa. Research work is in progress on improving SUNSAT's 15m multi-spectral imager to a 2.5m-resolution multi-sensor imager. This will require the use of Forward Motion Compensation for exposure control. This thesis presents Automatic Forward Motion Compensation for Micro Satellites using a new earth sensor that measures the bore-sight projection's speed on the earth directly. This sensor will have no drift and will make use of a series of images and crosscorrelation of them. The high-resolution imager's bore-sight motion can be controlled by this technique to ensure high quality stereo images. A control system based on the DLR-TUBSAT was designed and is simulated in MATLAB. This technique of measuring the bore-sight projection's speed on the earth directly was tested on a 2-axis telescope and used to measure random movement of a satellite. / AFRIKAANSE OPSOMMING: SUNSAT, Suid Afrika se eerste satelliet was vir 23 maande in werking in 'n wentelbaan na sy NASA-geborgde lasering op 23 Februarie 1999. SUNSAT is 'n satelliet wat ontwikkel is deur nagraadse studente aan die Universiteit van Stellenbosch. Navorsingswerk is aan die gang om SUNSAT se 15m resolusie multi-spektrale kamera te verbeter tot 'n 2.5m resolusie multi-sensor kamera. Die navorsing sal die gebruik van Voorwaartse Bewegingskompensasie benodig. Die tesis handeloor automatiese voorwaartse bewegingskompensasie vir mikrosatelliete deur gebruik te maak van 'n nuwe aardgerigte sensor, wat die siglyn projeksie se snelheid op die aarde direk sal meet. Die sensor sal gebruik maak van 'n reeks foto's wat gekruiskorreleer word en sal geen tempo-wegdrywing hê nie. Hoë resolusie kameras se siglyn kan deur middel van hierdie tegniek beheer word om hoë kwaliteit stereo foto's te verseker. 'n Beheerstelsel wat gebaseer is op dié van DLR-TUBSAT, is ontwerp en is gesimuleer in MATLAB. Die tegniek om die siglyn projeksie se snelheid op die aarde direk te meet, is getoets op 'n 2-as teleskoop en is gebruik om onverwille keurige beweging van 'n satelliet te meet.

No description available.

Made available in DSpace on 2014-11-10T11:09:41Z (GMT). No. of bitstreams: 0 Previous issue date: 2014-02-24Bitstream added on 2014-11-10T11:58:47Z : No. of bitstreams: 1 000790764.pdf: 3238994 bytes, checksum: f2fdaf11b3e4db2185fb27b148d75971 (MD5) / O processo de restauração de imagens tem como objetivo melhorar as imagens corrompidas por ruídos e borramentos. Técnicas iterativas podem controlar melhor o algoritmo de restauração a fim de restaurar detalhes em regiões borradas sem aumentar o ruído. Técnicas baseadas em Projeção em Conjuntos Convexos (Projections onto Convex Sets - POCS) tem sido utilizadas no contexto de restauração de imagens, projetando a solução em um hiperespaço até que algum critério de convergência seja encontrado. O resultado esperado é uma imagem melhor ao final de um número desconhecido de projeções. O número de conjuntos convexos e suas combinações permitem construir vários algoritmos de restauração de imagens baseados em POCS. O presente trabalho utiliza dois conjuntos convexos: Row-Action Projections (RAP) e Limited Amplitude (LA). O algoritmo RAP possui um parâmetro de relaxação que depende das características da imagem que será restaurada. Assim, valores errados de podem conduzir a uma pobre restauração. Propomos achar o valor de modelando o problema como de otimização e utilizando diferentes técnicas evolucionistas. Além disso, é possível utilizar os parâmetros de restauração aprendidas sobre uma imagem e empregá-los para uma outra imagem / The process of image restoration aims to enhance images corrupted by noise and blurred. Iterative techniques can better control the restoration algorithm in order to restore blurred regions in details without increasing noise. Techniques based on Projection Sets in Convex (Projections onto Convex Sets - POCS) have been used in the context of image restoration by projecting the solution in a hyperspace until some convergence criterion is met. The expected result is a better picture at the end of an unknown number of projections. The number of convex sets and its combinations allow you to build several image restoration algorithms based on POCS. This study uses two convex sets: Row Action Projections (RAP) and Limited Amplitude (LA). The RAP algorithm has a relaxation parameter depends on the characteristics of the image that will be restored. Thus, erroneous values of can lead to a poor restoration. We propose to find the value of as the problem of modeling and optimization using different evolutionary techniques. Furthermore, is possible to use the parameters learned in restoring an image, and use them to another image

Orientador: João Paulo Papa / Banca: Antonio Carlos Sementille / Banca: Alexandre Luís Magalhães Levada / Resumo: O processo de restauração de imagens tem como objetivo melhorar as imagens corrompidas por ruídos e borramentos. Técnicas iterativas podem controlar melhor o algoritmo de restauração a fim de restaurar detalhes em regiões borradas sem aumentar o ruído. Técnicas baseadas em Projeção em Conjuntos Convexos (Projections onto Convex Sets - POCS) tem sido utilizadas no contexto de restauração de imagens, projetando a solução em um hiperespaço até que algum critério de convergência seja encontrado. O resultado esperado é uma imagem melhor ao final de um número desconhecido de projeções. O número de conjuntos convexos e suas combinações permitem construir vários algoritmos de restauração de imagens baseados em POCS. O presente trabalho utiliza dois conjuntos convexos: Row-Action Projections (RAP) e Limited Amplitude (LA). O algoritmo RAP possui um parâmetro de relaxação ���� que depende das características da imagem que será restaurada. Assim, valores errados de ���� podem conduzir a uma pobre restauração. Propomos achar o valor de ���� modelando o problema como de otimização e utilizando diferentes técnicas evolucionistas. Além disso, é possível utilizar os parâmetros de restauração aprendidas sobre uma imagem e empregá-los para uma outra imagem / Abstract: The process of image restoration aims to enhance images corrupted by noise and blurred. Iterative techniques can better control the restoration algorithm in order to restore blurred regions in details without increasing noise. Techniques based on Projection Sets in Convex (Projections onto Convex Sets - POCS) have been used in the context of image restoration by projecting the solution in a hyperspace until some convergence criterion is met. The expected result is a better picture at the end of an unknown number of projections. The number of convex sets and its combinations allow you to build several image restoration algorithms based on POCS. This study uses two convex sets: Row Action Projections (RAP) and Limited Amplitude (LA). The RAP algorithm has a relaxation parameter ���� depends on the characteristics of the image that will be restored. Thus, erroneous values of ���� can lead to a poor restoration. We propose to find the value of ���� as the problem of modeling and optimization using different evolutionary techniques. Furthermore, is possible to use the parameters learned in restoring an image, and use them to another image / Mestre

A DISSERTATION PRESENTED TO THE FACULTY OF INFORMATICS AND DESIGN OF THE CAPE PENINSULA UNIVERSITY OF TECHNOLOGY IN FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MAGISTER TECHNOLOGIAE PHOTOGRAPHY CAPE PENINSULA UNIVERSITY OF TECHNOLOGY 2013 / This study deals with the pitfalls and standards associated with the digitisation of photographic artefacts in formal collections. The popularity of the digital medium caused a rapid increase in the demand for converting images into digital files. The need for equipment capable of executing the task successfully, the pressure on collection managers to display their collections to the world and the demand for knowledge needed by managers and operators created pressure to perform optimally and often in great haste. As a result of the rush to create digital image files to be displayed and to be preserved, the decisions that are being made may be questionable. The best choice of file formats for longevity, setting and maintaining standards to guarantee quality digital files and consultation with experts in the field of digitisation as well as attention to best practices are important aspects which must be considered. In order to determine the state of affairs in countries with an advanced knowledge and experience in the field of digitisation, a comprehensive literature study was done. It was found that enough information exists to enable collection managers in South Africa to make well informed decisions to ensure a high quality of digital collection. By means of questionnaires, a survey was undertaken amongst selected Western Cape image preservation institutions to determine the level of knowledge of the managers who are required to make informed decisions. The questionnaire was designed to give insight into choices being made regarding the technical quality, workflow and best practice aspects of digitisation. Comparing the outcome of the questionnaires with best practices and recommended standards in countries with an advanced level of experience it was found that not enough of this experience and knowledge is used by local collection managers although readily available. In some cases standards are disregarded completely. The study also investigated by means of questionnaires the perception of the digital preservation of image files by fulltime photographic students and volunteer members of the Photographic Society of South Africa. It was found that uncertainty exist within both groups with regard to file longevity and access to files in five to ten year's time. Digitisation standards are set and maintained by the use of specially designed targets which enable digitising managers to maintain control over the quality of the digital content as well as monitoring of equipment performance. The use of these targets to set standards were investigated and found to be an accurate and easy method of maintaining control over the standard and quality of digital files. Suppliers of digitising equipment very often market their equipment as being of a high quality and being able to fulfil the required digitisation tasks. Testing selected digitising equipment by means of specially designed targets proved however that potential buyers of equipment in the high cost range should be very cautious about suppliers' claims without proof of performance. Using targets to verify performance should be a routine check before any purchase. The study concludes with recommendations of implementing standards and it points to potential future research.

Thesis (MTech (Information Technology))--Peninsula Technikon, 2004 / Encryption was used whenever someone wanted to send a secret message to someone. The quality of the algorithm and key combination were the factors that ensured the strength of the system. However, until there were some automation one could not use complex methods for encryption because it simply took too long to encrypt and decrypt messages (even worse for images), manually. Optical technologies have recently been employed in encryption. Compared with traditional computer and electrical systems, optical technologies offer primarily two types of benefits, namely optical systems have an inherent capability for parallel processing, that is, rapid transmission of information, and information can be hidden in any of several dimensions, such as phase or spatial frequency. Optical systems therefore have an excellent capability for encoding information. In this project an image encryption technique was developed using exclusive-OR (XOR) operations in the optical domain before the captured image entered a digital computer network for further processing. A gray-level image of the object to be encrypted was converted a binary format and a pixel by pixel exclusive OR operation was performed on it with a random key-bit by making use of polarization encoding in LCD technology, before the encrypted binary image was detected by a CCD. The image may also be used as an input to a computer for comparison with a database. However, noise alters the encrypted bit and the value of the noisy encrypted bit is no longer binary. In this research the encryption techniques was evaluated and tested for applicability to encrypt and decrypt successfully. The noise performance was tested and compared. The technique was applied for image transmission over a wireless IP broadband link. Images (optical and thermal) captured from telemedicine application was transmitted and decrypted with reasonable success in this application.

O ato de executar zoom em imagens é uma tarefa que se aplica em diversas áreas que podem variar desde entretenimento até aplicações científicas. Um dos grandes desafios na área é manter a definição das bordas dos objetos da imagem sem que haja a criação de artefatos tais como aspecto serrilhado ou borramento. Diversos métodos de preservação de borda foram apresentados na literatura. Este trabalho apresenta a proposta de uma nova técnica de interpolação de imagens baseada em clustering que tem como objetivo aumentar a resolução da imagem em tons de cinza preservando as bordas dos objetos nela presentes com um método mais simples e de fácil implementação. Foram realizados testes da técnica proposta com diversas imagens de natureza diferente e seus resultados comparados aos métodos clássicos de interpolação de imagem encontrados na literatura. Para teste da eficácia foram consideradas a medida do PSNR e Correlação Cruzada com cada método comparado. Os resultados obtidos mostraram que a técnica é promissora e que cumpre os objetivos do projeto. / Image zooming is a task applicable to many areas which can vary from entertainment to scientific applications. A big challenge is image edge preserving without creating artifacts like blurring or blocking. Several methods for edge preserving were proposed in literature. This work presents a new technique proposal based on clustering which aims to increase gray scale image resolution preserving objects edges with a simple method and easy to implement. Many different types of images were used to make tests of the proposed technique and results are compared to classical methods of image interpolation found in literature. PSNR and Cross-Correlation measurements were used to compare efficiency between methods. Results showed that the technique is quite competitive and meets the project goals.