Development of new digital image analysis methods for determination of the geometrical properties of aggregates

Shape, angularity, and texture are the key geometrical properties which are frequently used to characterize aggregates. These properties have significant effects on the performance of construction materials such as asphalt and concrete. Also, there is no established and accepted general test method for determination of the geometrical properties of aggregates. Current standard test methods are subjective, labor-intensive and time consuming. In the absence of an effective and objective test method for quantifying aggregate geometrical properties, there is an urgent need for determining the geometrical properties of aggregates directly.The present dissertation focuses on to development of new modules for accurate determining of geometrical properties of aggregates. Three different modules; illuminated panel, laser based imaging, and virtual imaging systems have been developed for the several shape characterization of the aggregates using various digital image analysis techniques. The developed systems have been optimized to minimize the mistakes during image capturing and processing. Aggreganalysis software tool with user friendly surface property which was coded in C++ programming language allows accurate determination of geometric properties of aggregates such as angularity, compactness, concavity and eccentricity. Illuminated panel system may be implemented in 2D and in 3D image analysis of aggregates including depth information. For the one in 3D, MatLab® Image processing toolbox can be used to classify the aggregates according to their geometrical properties. Laser based imaging system enables capturing aggregates in 3D form. A special software surface was coded in MatLab® to recognize the geometrical properties of the aggregates. The developed virtual imaging system has advantages comparatively in capturing images in 3D with only one camcorder and with 4 special mirrors. The most important advantage of this economical virtual imaging system is to capture the real aggregate images in 3D with only one camcorder from 5 different surface angles. Finally, the magnitude of effects of the geometrical properties of aggregates has been determined on the mechanical properties of some mortar and concrete mixtures.