up with automated systems for assessment of road condition. For example, Haas et al (1997) developed an automated algorithm for detecting cracks and joints con- tion. Smith and Lin (1997) developed a fuzzy logic classification scheme for pavement distress condition. Oh et al (1997) developed iterative algorithm for overcoming noisy images of roads due to shadows and low light conditions. Koustsopoulos and Mishalani (1997) presented a model for distress assessment in a local (microscopic) and global (macroscopic) level using captured images of pavement. Lee (1993) presented a comparison between 15 different imaging al- rithms used in crack detection. Ground Penetration Radar (GPR) has also been used for pavement assessment. Special computer algorithms were developed for quick analysis of GPR data (Adeli & Hung 1993 and Maser 1996). Heiler and McNeil (1997) proposed a modified system for analyzing the GPR data using an artificial neural network (ANN). 2.3.2 Traffic Analysis and Control Currently imaging systems provide essential data for transportation and traffic engineering planning (Anon 1999). Machine vision techniques were introduced to intersection traffic signal control in the late 1970's (Chou and Sethi 1993). No- days, many systems have been developed all over the world for traffic analysis and control applications, in addition to image based systems for traffic violations. Nallamathu and Wang (1997) developed one of the first automated systems for license plate recognition using character recognition algorithm for the use in monitoring violators at toll stations and many other traffic applications.
The fundamental premise of this monograph is that transparent synthetic materials with geotechnical properties similar to those of natural soils can be used to study 3D deformation and flow problems in natural soils. Transparent soils can be made by matching the refractive index of synthetic soil materials and the pore fluid. This monographs presents the geotechnical behaviour of several families of transparent soils that can be combined to meet model-test requirements, in terms of strength, deformation, or permeability.
"Modelling with Transparent Soils" demonstrates how an optical system consisting of a laser light, a CCD camera, a frame grabber, and a PC can be used to measure spatial deformations in transparent soil models non-intrusively. Transparent soil models are sliced optically using a laser light sheet. A distinctive speckle pattern is generated by the interaction of the laser light and transparent soil. A 2D deformation field is obtained from two speckle images by using an image processing technique named adaptive cross-correlation, which is an advanced form of the digital image cross-correlation (DIC) algorithm that utilizes both window sizing and window shifting methods. The monograph demonstrates that comparison of 2D deformation fields between transparent soil and natural soil showed that the results were comparable in almost every aspect. Three dimensional fields can be produced by combining multiple 2D fields in Matlab.
Multiphase flow and surfactant flushing tests were also simulated using a layered transparent soil systems and several contaminants. The developed technology allows for visualizing the contamination concentration and evaluating the performance of remediation technologies in bench scale model tests.
