A inspeção visual de imagens radiográficas de soldas e tubulações é uma tarefa difícil para o ser humano, devido a sua característica repetitiva. Desta forma, a partir da digitalização dessas imagens radiográficas, é possível o desenvolvimento de ferramentas de inspeção automática, reduzindo os erros cometidos por inspeção visual. Este trabalho apresenta uma metodologia que tem o objetivo de automaticamente segmentar os defeitos em soldas e tubulações e extrair características que permitam o reconhecimento destes defeitos. Os resultados obtidos com o uso da metodologia desenvolvida neste trabalho mostram que é possível segmentar os defeitos em imagens radiográficas de soldas e levantar as características desses defeitos de forma automática, mesmo que tais defeitos sejam de difícil visualização. O trabalho também propõe o uso de parâmetros para calibrar o método proposto, sendo que tais parâmetros conferem uma maior maleabilidade ao método. Desta forma o método proposto para segmentar os defeitos foi testado com dois grupos de imagens, cada grupo proveniente de origem diferente, e obteve resultados satisfatórios conseguindo segmentar os defeitos em quase todas as imagens. Para extrair as características utilizou-se a rotulação das regiões segmentadas fechadas, este algoritmo coloca um label em cada possível defeito e assim permite levantar os seus parâmetros de forma automática. Por sua vez o método de segmentação de cordões de solda também se mostrou visualmente robusto conseguindo segmentar os cordões mesmo em imagens onde haviam outros objetos inseridos. / The visual inspection of radiographic images of welds and pipelines is a very repetitive task, very hard for humans. However, using digital radiographic images, it is possible to develop computational tools to be applied in the inspection process and to decrease the risks of inconsistent evaluations. This work presents a methodology to be applied in the automatic segmentation of defects in welds and pipes radiographic images, as well as to extract characteristics for the defects recognition. The tests and results, using the proposed methodology, show that it is possible to automatically perform the defect segmentation and characteristic extraction. As a result, some defects, difficult to be visually detected, were automatically segmented by the proposed method. Using some adjusting parameters, also proposed in this work, it is possible to adjust the method performance, obtaining more flexibility in the method applications. Two groups of images, originating from different sources, were tested and the results were satisfactory, since the defects, in almost all images, were segmented by the proposed method. A connected region labeling algorithm was used in order to make possible the automatic characteristics extraction for each detected defect. Besides, the proposed methodology shows to be robust also in some situations where other undesirable objects are presented in the image.

Thesis (MTech (Electrical Engineering))--Cape Peninsula University of Technology, 2005 / This dissertation is the reading and display ofDICOM medical images (Digital Imaging and Communication in Medicine) and production ofmodel artifacts of anatomical organs using Rapid Prototyping An algorithm to read these DICOM medical images was developed. It also displays pixel information ofthe image. When the DICOM image has been read and displayed, the information required to produce the anatomical artifact is extracted. These 2D slice images, MRI (Magnetic Resonance Imaging) and CT Scan (Computer Tomography) images are written to 3D file in SLC (Slice files) and STL (Stereolithography File Format) format. A 3D softcopy ofthe anatomical structure is created. At this stage, the clinician or surgeon can make any changes or require additional information to be added to the anatomical structure. With the 3D model available in STL format, a physical artifact is produced using Rapid Prototyping. The external edge ofthe anatomical structure can be produced using Rapid Prototyping as well as the outer rim with the internal structures. To produce the external surface ofthe structure, an outer rim edge detection algorithm has been developed. This will only extract the external surface ofthe structure. In addition to the softcopy ofthe structure, multiple organs can be displayed on the same image and this will give a representation ofthe interaction ofneighboring organs and structures. This is useful as both the normal anatomy as well as the infiltration ofthe abnormal pathology can be viewed simultaneously. One of the major limitations ofdisplaying the information in a 3D image is that the files are very large. Since 3D STL files use triangles to display the outer surface ofa structure, a method to reduce the file size and still keep the image information was developed. The triangle reduction method is a method to display the 3D information and to decrease the STL file size depending on the complexity ofthe outer surface ofthe structure. To ensure that the anatomical model s represented as in the DlCOM files, an Interpolation Algorithm was developed to reconstruct the outer ofthe model from 2D MRI or CT-Scan images. A word about computer models: Some of the programs and presentations are based on the real world. They model the real world and anatomical structures. It is very important to note that the models are created with software. Obviously a model is useful if it resembles reality closely, but it is only a prediction about the model itself. Models are useful because they help to explain why certain things happen and how interaction takes place. Models provide suggestions for how structures might look. Computer models provide answers very quickly. These are computer models representing the real structure. (Czes Kosniowski, 1983)

Thesis (MTech (Cartography))--Cape Peninsula University of Technology, 2016. / This case study investigates the change in the geographical boundaries by creating a Spatio-temporal mapping of Ciskei (one of the so-called Bantustans or Homelands) during the period of Apartheid. It examines the reasons for its establishment, and what impact the apartheid land legislation had on the geographical boundaries of Ciskei. GIS technology was used in this study to create time series animation and Static map to display the spatial change of the Ciskei boundaries. This investigation was split into quantitative and qualitative assessments. The aim of the quantitative assessments was to determine the amount of the spatial change of the Ciskei geographic boundary. The qualitative methods was used to investigate the map viewer’s understanding of the amount of the information in the static and animated maps. The results of qualitative assessments showed that static and animated maps have their respective advantages in the visualization of the map viewer. The importance of this research is to take advantage of time series mapping techniques to study the homeland areas in South Africa and see all the changes that have occurred as a result of a period of apartheid legislation. For this research, the following data were gathered: Attribute and metadata was the legislation and laws related to the land and the geographic data was the historical maps and coordinate data.

Panorama stitching increases the field of view in an image by assembling multiple views together. Traditional stitching techniques are proven to be effective only when dealing with parallax-free monocular images. Many challenges that remain unsolved in the stitching research area include how to stitch monocular images with large parallax, how to stitch stereoscopic images to maintain their stereoscopic consistency and original disparity distribution, and how to create panoramic videos with temporally coherent content. To provide more powerful stitching techniques with more universality, we first develop a parallax-tolerant image stitching technique. With the help of it, we then effectively extend the stitching techniques into the stereoscopic image and the video domain to assist users easily making stereoscopic panoramas and video panoramas. In this dissertation, we first introduce a parallax-tolerant stitching method, which is a local stitching method to stitch monocular images with large parallax. This method is based on the observation that input images do not need to be perfectly aligned over the whole overlapping region for stitching. Instead, they only need to be aligned in a way that there exists a local region where they can be seamlessly blended together. We develop a randomized algorithm to search for a local homography, which, combined with content-preserving warping, allows for optimal stitching. Our experiments show that our method can effectively stitch images with large parallax that are difficult for existing methods. After studying the problem of regular 2D image stitching, we continue to research 3D image stitching in this dissertation. In particular, we develop a technique for stitching stereoscopic panoramas from stereo images casually taken using a stereo camera. Stereoscopic image stitching needs to address three challenges: how to deal with parallax, how to stitch the left- and right-view panorama consistently, and how to take care of disparity during stitching. We address these challenges by first stitching the left images with the parallax-tolerant image stitching method to create an artifact-free left view panorama, then stitching the disparity maps with disparity optimization, finally warping and stitching the right images according to the stitched disparity map and the left view panorama. Experiment results show that our technique allows for easy production of high quality stereoscopic panoramas that deliver a pleasant stereoscopic 3D viewing experience. With the 3D image stitching problem addressed, we further study a more complex and challenging task of video stitching. We contribute two video stitching techniques, namely the motion map guided video stitching and the feature trajectory guided video stitching. Our techniques stitch pre-synchronized videos captured from a fixed or hand-held camera array which contains multiple cameras with fixed inter-camera configurations. One unique challenge for video stitching is how to maintain temporal coherence. To address this problem, we propose to consistently stitch frames with the guidance of the target camera motion path. In particular, we develop two techniques using dense motion maps and sparse motion vectors to compute the target camera motion path. Afterwards, we warp and stitch frames according to the target camera motion path to create panoramic videos with temporal coherence. Experiments show that our methods can improve the overall panoramic video stitching quality compared with existing methods.

The professional work of a radiographer encompasses both patient care and the use of technology. The technology employed could either be analogue or digital technology. Since 1973, the analogue imaging system has slowly been replaced by digital radiography imaging systems. Despite the many advantages of digital imaging it does present the radiographer with added responsibilities. Furthermore, analogue-trained radiographers have found adjusting to digital imaging especially challenging. The aim of the study was to explore and describe the experiences of analogue-trained radiographers utilising digital imaging in projection radiography with the intention of developing guidelines to equip radiography managers to assist analogue-trained radiographers to better utilise digital imaging. The researcher used Schlossberg’s Transition Theory as a lens to look at the experiences of analogue-trained radiographers using digital imaging to produce radiographs. The research study used a qualitative design which was explorative, descriptive and contextual in nature. The target population included all diagnostic radiographers (public and private) in the local municipality who were registered with the Health Professions Council of South Africa. Purposive sampling was employed to select the radiographers that represented all radiographers in the Nelson Mandela Bay Health District. The sample included all radiographers who fulfilled the identified selection criteria. The selected participants were recruited to take part in in-depth, semi-structured individual interviews. The data was analysed using a computer-aided qualitative data analysis software package, ATLAS.ti. The trustworthiness of this study was ensured by applying Guba’s model of trustworthiness that includes credibility, transferability, dependability and confirmability. The ethical principles of respect for persons, beneficence and justice, as espoused by the Belmont Report, were adhered to in order to ensure that the study was conducted in an ethical manner. Two themes emanated from the data, namely the evolution of the radiographer when faced with the advances in technology as well the role that the work environment played in the manner that the participants experienced the change. The experiences of the participants were described using direct quotations from the interviews and a literature control was used to verify the participants’ experiences. Evidence was found of radiographer indifference towards exposure selection, dose optimisation and placement of anatomical side markers when utilising digital imaging. Finally, guidelines were developed to equip radiography managers to assist analogue-trained radiographers to better utilise digital imaging. In addition, the guidelines will assist all other radiographers to better utilise digital imaging.

This thesis investigates the viability of applying a shape-from-shading technique to SAR imagery. A shape-from-shading algorithm is derived and tested on a single site for which both a Seasat SAR image and Digitial Elevation Model (DEM) were available. The shape-from-shading technique used in this thesis follows an approach proposed by Frankot and Chellappa for processing slant range SAR imagery. The algorithm incorporates a one-step technique for projecting non-integrable surface orientation estimates onto an integrable set in the frequency domain along with the iterative convergent shape-from-shading algorithm of Brooks and Horn. The significant issues and choices made in implementing the shape-from-shading algorithm and in preparing the SAR data and DEM are discussed. The shape-from-shading algorithm was applied to both the test site SAR image and images synthesized from the DEM. Reflectance models were derived from the SAR image and DEM. By quantitatively comparing the shape-from-shading results with the initial conditions used for the experiments, it was found that the algorithm produced substantially better results when applied to the synthesized images; however, when applied to the SAR image, there was no significant improvement over the initial conditions. / Science, Faculty of / Computer Science, Department of / Graduate

A system has been implemented which is able to detect symmetrical groupings in edge images. The initial stages of the algorithm consist of edge detection, curve smoothing, and the extension of the perceptual grouping phase of the SCERPO [Low87] vision system to enable detection of instances of endpoint proximity and curvilinearity among curved segments. The symmetry detection stage begins by first locating points along object boundaries which are significant in terms of curvature. These key points are then tested against each other in order to detect locally symmetric pairs. An iterative grouping procedure is then applied which matches these pairs together using a more global definition of symmetry. The end result of this process is a set of pairs of key points along the boundary of an object which are bilaterally symmetric, along with the axis of symmetry for the object or sub-object. This paper describes the implementation of this system and presents several examples of the results obtained using real images. The output of the system is intended for use as indexing features in a model-based object recognition system, such as SCERPO, which requires as input a set of spatial correspondences between image features and model features. / Science, Faculty of / Computer Science, Department of / Graduate

Colour images can be formed by the combination of stimuli in three primary colours. As a result, digital colour images are typically represented as a triplet of values, each value corresponding to the stimulus of a primary colour. The precise stimulus that the eye receives as a result of any particular triplet of values depends on the display device or medium used. Photographic film is one such medium for the display of colour images. This work implements a software system to calibrate the response given to a triplet of values by an arbitrary combination of film recorder and film, in terms of a measurable film property. The implemented system determines the inverse of the film process numerically. It is applied to calibrate the Optronics C-4500 colour film writer of the UBC Laboratory for Computational Vision. Experimental results are described and compared in order to estimate the expected accuracy that can be obtained with this device using commercially available film processing. / Science, Faculty of / Computer Science, Department of / Graduate

This thesis addresses the problem of computationally estimating the motion of pack ice in sequential digital satellite images. The problem is posed in terms of linear filter theory and is solved by minimizing the error variance. The intuitive use of cross correlation and edge detection are shown to flow naturally from this approach. The theoretical framework also allows a geometric intuition into the action of the filter which is not possible through ad hoc methods. The noise corrupting the filtering process is investigated and the filter is implemented through both a first order method common to image processing, and a more sophisticated second order approach from computational vision. The class of imagery for which the filtering system is appropriate is discussed and the images chosen for the experiments are shown to be representative of this class. The experimental results reveal the power of the system in estimating ice motion, and some analysis of the derived motion is performed by comparison to a simple theory of wind-driven ice motion. The failings of the system are discussed and improvements are suggested. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

An adaptive image compression coding technique, ACC, is presented. This algorithm is shown to preserve edges and give better quality decompressed pictures and better compression ratios than that of the Absolute Moment Block Truncation Coding. Lookup tables are used to achieve better compression rates without affecting the visual quality of the reconstructed image. Regions with approximately uniform intensities are successfully detected by using the range and these regions are approximated by their average. This procedure leads to further reduction in the compression data rates. A method for preserving edges is introduced. It is shown that as more details are preserved around edges the pictorial results improve dramatically. The ragged appearance of the edges in AMBTC is reduced or eliminated, leading to images far superior than those of AMBTC. For most of the images ACC yields Root Mean Square Error smaller than that obtained by AMBTC. Decompression time is shown to be comparable to that of AMBTC for low threshold values and becomes significantly lower as the compression rate becomes smaller. An adaptive filter is introduced which helps recover lost texture at very low compression rates (0.8 to 0.6 b/p, depending on the degree of texture in the image). This algorithm is easy to implement since no special hardware is needed. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate