Scale space analysis and neural arbitration for 3-dimensional edge detection

In the past much research has been carried out into the development of precise and fast processing techniques for object detection within images. However, these have not addressed the question of variable scale. Multiscale analysis techniques have attempted to address this problem of variable scale. The work so far reported has focused on region extraction and edge detection of 2-dimensional objects. Here the noise and illumination effects on the images are less than would be found in the case of a 3-dimensional object. The quality of the detected edges for both high and low-contrast images, was the main emphasis of these previous studies. In this paper both the quality of the detected edges and the introduction of the noise and illumination effects due to the third dimension are considered. The use of neural networks for edge detection is in its infancy, however, the initial results are very promising. They have not as yet been applied in multiscale analysis. Investigations are reported into the use of scale space analysis for 3-dimensional object recognition. These results then form the basis for the use of neural network arbitration of the fast Laplacian of the Gaussian operator to carry out automatic recognition