Estrategias para identificar oclusiones y planificación monocular para una mejora de la percepción visual de la escena

Esta Tesis doctoral está orientada al estudio de estrategias y técnicas para el tratamiento de oclusiones. Las oclusiones suponen uno de los principales problemas en la percepción de una escena mediante visión por computador. Las condiciones de luz, los puntos de vista con los que se captura información de la escena, las posiciones y orientaciones de los objetos presentes en la escena son algunas de las causas que provocan que los objetos puedan quedar ocluidos parcialmente. Las investigaciones expuestas en esta Tesis se pueden agrupar en función de su objetivo en dos grupos: técnicas cuya finalidad es detectar la presencia de oclusiones y estrategias que permiten mejorar la percepción de un sistema de visión por computador, aun en el caso de la presencia de oclusiones. En primer lugar, se han desarrollado una serie de técnicas orientadas a la detección de oclusiones a partir de procesos de extracción de características y de segmentación color en imágenes. Estas técnicas permiten definir qué regiones en la imagen son susceptibles de considerarse zonas de oclusión, debido a una mala percepción de la escena, como consecuencia de observarla con un mal punto de vista. Como aplicación de estas técnicas se han desarrollado algoritmos basados en la segmentación color de la imagen y en la detección de discontinuidades mediante luz estructurada. Estos algoritmos se caracterizan por no incluir conocimiento previo de la escena. En segundo lugar, se han presentado una serie de estrategias que permiten corregir y/o modificar el punto de vista de la cámara con la que se observa la escena. De esta manera, las oclusiones identificadas, mediante los métodos expuestos en la primera parte de la Tesis, y que generalmente son debidas a una mala localización de la cámara pueden ser eliminadas o atenuadas cambiando el punto de vista con el que se produce la observación. En esta misma línea se presentan dos estrategias para mejorar la posición y orientación espacial de la cámara cuando ésta se emplea para la captura de imágenes en procesos de reconocimiento. La primera de ellas se basa en la retroproyección de características obtenidas de una imagen real, a partir de una posición cualquiera, en imágenes virtuales correspondientes a las posibles posiciones que puede adoptar la cámara. Este algoritmo lleva a cabo la evaluación de un mapa de distancias entre estas características buscando en todo momento, maximizar estas distancias para garantizar un mejor punto de vista. La ventaja radica en que en ningún caso se hace necesario mover la cámara para determinar una nueva posición que mejore la percepción de la escena. La segunda de estas estrategias, busca corregir la posición de la cámara buscando la ortogonalidad. En este caso, se ha partido de la hipótesis inicial de que la mayor superficie visible siempre se suele conseguir situando la cámara ortogonalmente al plano en el que se sitúa el objeto.

The blindness of Dr. Gray, or, The final law /

Mode of access: Internet.

Deaf-Blindness : national organizations and resources /

"January 1993."

Building a library collection on blindness and physical disabilities : basic materials and resources.

Mode of access: Internet.

Building a library collection on blindness and physical handicaps : basic materials and resources.

Mode of access: Internet.

Building a library collection on blindness and physical handicaps : basic materials and resource.

Mode of access: Internet.

Prevention of Blindness

Mode of access: Internet.

The blindness of virtue,

Drama.

The physiology of vision with special reference to colour blindness.

Mode of access: Internet.

Does interhemispheric competition mediate motion-induced blindness? A transcranial magnetic stimulation study

Motion-induced blindness (MIB) is a phenomenon, perhaps related to perceptual rivalry, where stationary targets disappear and reappear in a cyclic mode when viewed against a background (mask) of coherent, apparent 3-D motion. Since MIB has recently been shown to share similar temporal properties with binocular rivalry, we probed the appearance-disappearance cycle of MIB using unilateral, single-pulse transcranial magnetic stimulation (TMS)-a manipulation that has previously been shown to influence binocular rivalry. Effects were seen for both hemispheres when the timing of TMS was determined prospectively on the basis of a given subject's appearance-disappearance cycle, so that it occurred on average around 300 ms before the time of perceptual switch. Magnetic stimulation of either hemisphere shortened the time to switch from appearance to disappearance and vice versa. However, TMS of left posterior parietal cortex more selectively shortened the disappearance time of the targets if delivered in phase with the disappearance cycle, but lengthened it if TMS was delivered in the appearance phase after the perceptual switch. Opposite effects were seen in the right hemisphere, although less marked than the left-hemisphere effects. As well as sharing temporal characteristics with binocular rivalry, MIB therefore seems to share a similar underlying mechanism of interhemispheric modulation. Interhemispheric switching may thus provide a common temporal framework for uniting the diverse, multilevel phenomena of perceptual rivalry.

Amodal completion with background determines depth from monocular gap stereopsis

Grove, Gillam, and Ono [Grove, P. M., Gillam, B. J., & Ono, H. (2002). Content and context. of monocular regions determine perceived depth in random dot, unpaired background and phantom stereograms. Vision Research, 42, 1859-1870] reported that perceived depth in monocular gap stereograms [Gillam, B. J., Blackburn, S., & Nakayama, K. (1999). Stereopsis based on monocular gaps: Metrical encoding of depth and slant without matching contours. Vision Research, 39, 493-502] was attenuated when the color/texture in the monocular gap did not match the background. It appears that continuation of the gap with the background constitutes an important component of the stimulus conditions that allow a monocular gap in an otherwise binocular surface to be responded to as a depth step. In this report we tested this view using the conventional monocular gap stimulus of two identical grey rectangles separated by a gap in one eye but abutting to form a solid grey rectangle in the other. We compared depth seen at the gap for this stimulus with stimuli that were identical except for two additional small black squares placed at the ends of the gap. If the squares were placed stereoscopically behind the rectangle/gap configuration (appearing on the background) they interfered with the perceived depth at the gap. However when they were placed in front of the configuration this attenuation disappeared. The gap and the background were able under these conditions to complete amodally. (c) 2006 Elsevier Ltd. All rights reserved.

Contrast discrimination with simultaneous monocular and dichoptic masks

In experiments reported elsewhere at this conference, we have revealed two striking results concerning binocular interactions in a masking paradigm. First, at low mask contrasts, a dichoptic masking grating produces a small facilitatory effect on the detection of a similar test grating. Second, the psychometric slope for dichoptic masking starts high (Weibull ß~4) at detection threshold, becomes low (ß~1.2) in the facilitatory region, and then unusually steep at high mask contrasts (ß~5.5). Neither of these results is consistent with Legge's (1984 Vision Research 24 385 - 394) model of binocular summation, but they are predicted by a two-stage gain control model in which interocular suppression precedes binocular summation. Here, we pose a further challenge for this model by using a 'twin-mask' paradigm (cf Foley, 1994 Journal of the Optical Society of America A 11 1710 - 1719). In 2AFC experiments, observers detected a patch of grating (1 cycle deg-1, 200 ms) presented to one eye in the presence of a pedestal in the same eye and a spatially identical mask in the other eye. The pedestal and mask contrasts varied independently, producing a two-dimensional masking space in which the orthogonal axes (10X10 contrasts) represent conventional dichoptic and monocular masking. The resulting surface (100 thresholds) confirmed and extended the observations above, and fixed the six parameters in the model, which fitted the data well. With no adjustment of parameters, the model described performance in a further experiment where mask and test were presented to both eyes. Moreover, in both model and data, binocular summation was greater than a factor of v2 at detection threshold. We conclude that this two-stage nonlinear model, with interocular suppression, gives a good account of early binocular processes in the perception of contrast. [Supported by EPSRC Grant Reference: GR/S74515/01]