Speckle Interferometry Information Extraction Technology And Application

Electronic speckle pattern interferometry (ESPI) is a whole-field, nondestructive measurement technique, which is well-known for deformative and nondestructive measurement of coarse surface. As using this technique, accurate extraction of phase value is of fundamental importance for the successful measurement of displacement, strain and vibration. At present, two of the most comprehensive phase extraction techniques are the fringe-centerline method and the phase-shifting technology. The latter method is used for static measurement mainly, and is paid more attention because the whole-field phase information is available.As using fringe-centerline method to extract phase, the strong grain-shape random noise leads to heavy restrain to the fringe on visibility and resolving ability. It is difficult of accurate extraction of phase value from fringe patterns. Therefore, research on effectively filter to the fringe patterns is of fundamental importance for the development of ESPI. The partial differential equations (PDE) filter methods and the variational filter methods were used for noise reduction. We discussed the filter effect of various PDE filter models and variational filter models to ESPI fringe patterns, analyzed reasons of success or failure, and proposed more effective PDE filter models based on fringe features. After denoising, a series of operations, binarization, skeleton extraction, fringe order assignation and interpolation, were implemented and the three-dimension phase map was obtained. For phase-shifting technology, the generation modes for phase map, filter method for wrapped phase map and the phase unwrapped technique were studied.Owing to the prominent advantages, such as real-time display, high sensitivity and whole-field measure, ESPI is widely used in industrial nondestructive detection. Applying the electronic shearing speckle pattern interferometry (ESSPI) and the phase-shifting technology, a miniaturization shearing phase-shifting nondestructive detection system was set up for bubbles detection of tire.The major innovations achieved in the dissertation are as follows:(1) For noise reduction of ESPI fringe patterns, two variational filter models, i.e. PM variational model and total variational model, were improved.(2) Two new PDE filter models were presented. Considering the imfromation of fringe density and the feature of containing multiplicative noise, a homomorphic PDE filter method based on fringe orientation was proposed. And combining the isophote line and the normal curve, a new anisotropic PDE filter method was presented. The filter effect has been improved distinctly.(3) For phase extraction technology, the MBO algorithm based on the heat conduction equation was introduced for the binarization of fringe patterns for the first time, and the binarization effect by our improved algothm is better than that by those traditional threshold methods. Furthermore, the least-squares phase unwrapped method was reformed and the precision was improved.(4) A miniaturization shearing speckle phase-shifting system for bubbles detection of tire was set up. The overall design was made for the system and some images about tire defect were acquired. The system can not be quakeproof. The feasibility of the system was validated.