Modelling and simulating affine-style cameras : an investigation into the use of thin-film directional filters in large-area cameras

Abstract

This thesis examines the behaviour of a large area thin-film camera. A new camera design, termed affine-style, was proposed which uses a thin-film directional filter to collimate light. This design had the benefits of being extremely scalable to different surface sizes and extremely thin (potentially several microns) allowing it to be integrated into surfaces such as walls or chassis. To determine the behaviour of this camera a novel model was established termed pseudo-affine. This camera was successfully simulated using a ray tracer and a realistic directional filter model. Whilst the simulation demonstrated such a camera is functional, it also showed that with current technology it is limited to short range usage, such as indoors, capturing large scale features, such as silhouettes of objects or humans, due to extreme blurriness. To achieve these discoveries this research developed the pseudo-affine camera model using the affine transform as a basis. Using this new model, the key controlling variables were identified, these determined the camera’s intrinsic properties such as the size of the sensor of the camera and its ability to collimate light, which in this project was termed affineness. Using these variables and the pseudo-affine theory a ray tracer model was developed and tested against a Söller collimation camera. Using the ray tracer model, a series of simulations further investigated the effects of the camera’s intrinsic and extrinsic variables. This showed that affine-style camera’s output quality is governed by the ratio between it Field-of-View size and the size of the feature being captured. The larger the ratio, the more the image became blurred, and the worse image quality became. The relationship between this ratio and image quality was calculated and was used as the foundation of a novel design methodology for the affine-style camera.