Note:

No description available.

No description available.

An important problem for the human factors specialist involves specifying the extent to which the quality of a displayed image will affect observer performance. One approach to this problem has been to develop metrics of image quality and to correlate measured image quality with observer performance. Of the metrics that have been developed, the Modulation Transfer Factor Area (MTFA) has proven to be a useful predictor of observer performance. The MTFA metric takes into account the system's Modulation Transfer Function and observer threshold curves. Two experiments were performed to assess the effects of the quality of a televised image on eye movements and search-related dependent measures. The first experiment search task involved having subjects perform an air-to-ground search during simulated flight. The quality of the image presented was varied by either passing, low-pass filtering, or attenuating the video signal and by adding electrical white noise to the video signal. The results of this experiment indicate that (1) at the highest level of electrical noise added, the percent of correct target acquisitions was decreased moderately, (2) the larger the target, the higher the percent correct responses, (3) the low-pass filtering of the video signal led to shorter ground ranges at acquisition for the large-sized targets, and (4) that the larger the target, the longer the fixation duration. Low to moderate correlations between MTFA and performance measures generally indicated that as MTFA increases performance improves, and that as MTFA increases fixation duration decreases. The search task of the second experiment consisted of having the subjects search for a designated letter or numeral across a televised picture of randomly positioned letters and numerals. The quality of the picture was varied by either passing, low-pass filtering, high-pass filtering, or attenuating the video signal and by adding electrical white noise to the video signal. The results of this experiment indicated that (1) the high-pass filtered high noise level condition led to significantly longer search times; and (2) the fixation times associated with the high-pass filtered condition were longer than those associated with the low-pass filtered, at- tenuated, and unfiltered unattenuated conditions, and that this effect was most pronounced under high noise level conditions. Correlations between MTFA and performance measures indicated that increases in MTFA lead to decreases in search time and decreases in fixation duration. The results of tests among fixation and saccade duration distribution across both experiments consistently showed significant individual differences and non-normality. Examination of pooled fixation duration distributions showed large dispersion differences across the static and dynamic imagery experiments. This result, and the differences of pooled fixation and saccade duration distribution ranges do not support the search model assumption of constant glimpse durations and reveal serious.flaws in the structure of search models. / Ph. D.

To simulate the edge perception ability of human eyes and detect scene edges from an image, context information and world constraints must be employed in the edge detection process. To accomplish this, two Bayesian decision theoretic frameworks for context dependent edge detection are developed around the local facet edge detector. The first approach uses all the context in the neighborhood of a pixel. The second approach uses the context of the whole image. The mechanism of the context edge detector then assigns a pixel the most probable edge state which is consistent with its assumed edge context. We also demonstrate how world constraints can aid the edge detection process with a lighting compensation and a curvature constraint scheme. The context information and world constraints can also be used to evaluate the performance of different edge detectors. A general edge coherence measure, a robust edge thinness measure, and a general edge correctness measure are developed. Upon comparing the performance of the edge detectors with the context free second directional derivative zero-crossing edge operator, we find that the context dependent edge detector is superior; the world constrained context free edge detectors can also improve the edge result. Finally, some simple edge detection schemes based on morphologic operations are discussed and evaluated. Although their performances are not as good as the other edge operators described in this dissertation, they are acceptable in the images which have reasonably high signal-to-noise ratio. These morphological edge operations can be realized most efficiently in machine vision systems that have special hardware designed for morphologic operations. / Ph. D.

Thesis (MSc)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: In this thesiswe approach the problem of tracking multiple people individually in a video sequence. Automatic object detection and tracking is non-trivial as humans have complex and mostly unpredictable movements, and there are sensor noise and measurement uncertainties present. We consider traditional object detection methods and decide to use thermal data for the detection step. This choice is supported by the robustness of thermal data compared to colour data in unfavourable lighting conditions and in surveillance applications. A drawback of using thermal data is that we lose colour information, since the sensor interprets the heat emission of the body rather than visible light. We incorporate a colour sensor which is used to build features for each detected object. These features are used to help determine correspondences in detected objects over time. A problem with traditional blob detection algorithms, which typically consist of background subtraction followed by connected-component labelling, is that objects can appear to split or merge, or disappear in a few frames. We decide to add ‘dummy’ blobs in an effort to counteract these problems. We refrain from making any hard decisions with respect to the blob correspondences over time, and rather let the system decide which correspondences are more probable. Furthermore, we find that the traditional Markovian approach of determining correspondences between detected blobs in the current time step and only the previous time step can lead to unwanted behaviour. We rather consider a sequence of time steps and optimize the tracking across them. We build a composite correspondence model and weigh each correspondence according to similarity (correlation) in object features. All possible tracks are determined through this model and a likelihood is calculated for each. Using the best scoring tracks we then label all the detections and use this labelling as measurement input for a tracking filter. We find that the window tracking approach shows promise even though the data we us for testing is of poor quality and noisy. The system struggles with cluttered scenes and when a lot of dummy nodes are present. Nonetheless our findings act as a proof of concept and we discuss a few future improvements that can be considered. / AFRIKAANSE OPSOMMING: In hierdie tesis benader ons die probleemomverskeiemense individueel in ’n video-opname op te spoor en te volg. Outomatiese voorwerp-opsporing en -volging is nie-triviaal, want mense het komplekse en meestal onvoorspelbare bewegings, en daar is sensor-ruis en metingonsekerhede teenwoordig. Ons neem tradisionele voorwerp-opsporing metodes in ag en besluit om termiese data te gebruik vir die opsporingstap. Hierdie keuse word ondersteun deur die robuustheid van termiese data in vergelyking met kleur data in ongunstige lig-kondisies en in sekuriteitstoepassings. Die nadeel van die gebruik van termiese data is dat ons kleur inligting verloor, aangesien die sensor die hitte vrystelling van die liggaam interpreteer, eerder as sigbare lig. Ons inkorporeer ’n kleur-sensor wat gebruik word om die kenmerke van elke gevolgde voorwerp te bou. Hierdie kenmerke word gebruik om te help om ooreenkomste tussen opgespoorde voorwerpe te bepaal met die verloop van tyd. ’n Probleem met die tradisionele voorwerp-opsporing algoritmes, wat tipies bestaan uit agtergrond- aftrekking gevolg deur komponent-etikettering, is dat dit kan voorkom asof voorwerpe verdeel of saamsmelt, of verdwyn in ’n paar rame. Ons besluit om ‘flous’-voorwerpe by te voeg in ’n poging om hierdie probleme teen te werk. Ons weerhou om enige konkrete besluite oor opgespoorde voorwerpe se ooreenkomste met die verloop van tyd te maak, en laat die stelsel eerder toe om te besluit watter ooreenkomste meer waarskynlik is. Verder vind ons dat die tradisionele Markoviaanse benadering vir die bepaling van ooreenkomste tussen opgespoorde voorwerpe in die huidige tydstap en die vorige een kan lei tot ongewenste gedrag. Ons oorweeg eerder ’n reeks van tydstappe, of ’n venster, en optimeer die volg van voorwerpe oor hulle. Ons bou ’n saamgestelde ooreenstemmingsmodel en weeg elke ooreenstemming volgens die ooreenkoms (korrelasie) tussen voorwerpe se kenmerke. Alle moontlike spore word deur hierdie model bepaal en ’n waarskynlikheid word bereken vir elkeen. Die spore met die beste tellings word gebruik om al die opsporings te nommeer, en hierdie etikettering word gebruik as meting-inset vir ’n volgingsfilter. Ons vind dat die venster-volg benadering belowend vaar selfs al is die invoerdata in ons toetse van swak gehalte en ruiserig. Die stelsel sukkel met besige tonele en wanneer baie flous-voorwerpe teenwoordig is. Tog dien ons bevindinge as ’n bewys van konsep en ons bespreek ’n paar verbeterings wat in die toekoms oorweeg kan word.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2005. / The purpose of this study was to investigate multi-spectral tracking. Various algorithms were investigated and developed to enhance the contrast between target and non-target classes. Different tracking algorithms were implemented on the resulting grayscale input. A physical tracking system consisting of a video input processor and DSP was designed and built to implement algorithms and investigate the viability of realtime multi-spectral tracking. It is illustrated that conventional intensity tracking clouds the available information and that by studying various spectral inputs information is extracted more efficiently from the available data.

The use of computers for digital image recognition has become quite widespread. Applications include face recognition, handwriting interpretation and fmgerprint analysis. A feature vector whose dimension is much lower than the original image data is used to represent the image. This removes redundancy from the data and drastically cuts the computational cost of the classification stage. The most important criterion for the extracted features is that they must retain as much of the discriminatory information present in the original data. Feature extraction methods which have been used with neural networks are moment invariants, Zernike moments, Fourier descriptors, Gabor filters and wavelets. These together with the Neocognitron which incorporates feature extraction within a neural network architecture are described and two methods, Zernike moments and the Neocognitron are chosen to illustrate the role of feature extraction in image recognition. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 1996.

In the past few years, violence detection has become an increasingly rele- vant topic in computer vision with many proposed solutions by researchers. This thesis proposes a solution called Criminal Aggression Recognition Engine (CARE), an OpenCV based Java implementation of a violence detection system that can be trained with video datasets to classify action in a live feed as non-violent or violent. The algorithm extends existing work on fast ght detection by implementing violence detection of live video, in addition to prerecorded video. The results for violence detection in prerecorded videos are comparable to other popular detection systems and the results for live video are also very encouraging, making the work proposed in this thesis a solid foundation for improved live violence detection systems. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection

Eye fixations of the face are normally directed towards either the eyes or the mouth, however the proportions of gaze to either of these regions are dependent on context. Previous studies of gaze behavior demonstrate a tendency to stare into a target’s eyes, however no studies investigate the differences between when participants believe they are engaging in a live interaction compared to knowingly watching a pre-recorded video, a distinction that may contribute to studies of memory encoding. This study examined differences in fixation behavior for when participants falsely believed they were engaging in a real-time interaction over the internet (“Real-time stimulus”) compared to when they knew they were watching a pre-recorded video (“Pre-recorded stimulus”). Results indicated that participants fixated significantly longer towards the eyes for the pre-recorded stimulus than for the real-time stimulus, suggesting that previous studies which utilize pre-recorded videos may lack ecological validity. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2016. / FAU Electronic Theses and Dissertations Collection