The effect of nonlinearity in robotic vision systems

Abstract

In this thesis, an application of generic approach to nonlinear image processing is described which is implemented in real-life application. Consideration is given to the general geometrical resampling process in which output pixels are estimated by interpolation of input pixels. However, here the geometrical resampling process is not at regular time interval. This generic approach can be used to solve several kinds of nonlinear image restoration problems when the image is being grabbed by a line scan camera. With a little modification, this algorithm can be easily implemented in hardware, so that image can be corrected during the grabbing period.