Adaptive and quantitative stereoscopic image analysis for radiographic and aerial photographic applications

This dissertation research sought to integrate relevant fuzzy logic and quantitative stereoscopic methods to solve selected technical problems currently existing in two closely related fields: X-ray radiography and aerial photography.; The first goal of this research project was to evaluate the feasibility and limitations of applying fuzzy-based methods for pattern recognition; particularly for recognizing color objects in a complex background under varying illumination. The knowledge and experience gained through this study was combined with photogrammetric or stereoscopic methods for analyzing images obtained by aerial photography and X-ray radiography, respectively.; The second goal of this project was to determine the technical feasibility of a new technology, quantitative stereo radiographic image analysis (QS-RIA), for industrial non-destructive evaluation (NDE) and medical diagnostic applications. The work has included: (1) refinements and integration of a current theory on stereo radiography with relevant pattern recognition methods to establish a science basis for QS-RIA, (2) design, construction, and evaluation of a new QS-RIA apparatus for stereoscopically displaying radiographic images and quantitatively assessing the internal features of an object (e.g. defects inside a structure or anomalies inside a human body), and (3) identification of the advantages and limitations of the proposed technology. The technology offers an inexpensive method of permitting an X-ray radiographic apparatus operator to: (a) conveniently display a three-dimensional view of the internal structure of an object, (b) quantitatively evaluate the spatial relations among various internal features, and (c) rapidly determine the exact location (x, y, z) of a specific internal feature.; The third goal of this project was to provide an improved method of stereoscopically displaying optic pairs of aerial photographic images and allow for more convenient and rapid determination of the location of a feature image point. The developed quantitative stereo spectroscopic image analysis (QS-SIA) approach provides an improved method for stereoscopic viewing of paper- or plastic-based images as well as video or digital images. The method also permits quantitative determination of the depth or height of a feature identified in the images.