Image analysis, modeling, enhancement, restoration, feature extraction and their applications in nondestructive evaluation and radio astronomy

摘要

The principal topic of this dissertation is the development and application of signal and image processing to Nondestructive Evaluation (NDE) and radio astronomy;The dissertation consists of nine papers published or submitted for publication. Each of them has a specific and unique topic related to signal processing or image processing in NDE or radio astronomy. Those topics are listed in the following. (1) Time series analysis and modeling of Very Large Array (VLA) phase data. (2) Image analysis, feature extraction and various applied enhancement methods for industrial NDE X-ray radiographic images. (3) Enhancing NDE radiographic X-ray images by adaptive regional Kalman filtering. (4) Robotic image segmentation, modeling, and restoration with a rule based expert system. (5) Industrial NDE radiographic X-ray image modeling and Kalman filtering considering signal-dependent colored noise. (6) Computational study of Kalman filtering VLA phase data and its computational performance on a supercomputer. (7) A practical and fast maximum entropy deconvolution method for deblurring industrial NDE X-ray and infrared images. (8) Local feature enhancement of synthetic radio images by adaptive Kalman filtering. (9) A new technique for correcting phase data of a synthetic-aperture antenna array.