Determination of Elongated Aggregates through Computer Vision-Based Technique

The traditional measurement technique for finding elongated aggregates involves sieve analysis followed by passing each particle through an elongation gauge. In this study, machine learning has been used as a computer vision-based methodology to determine the elongation index of coarse aggregate particles. Images of aggregates were captured, and an algorithm was developed for contour detection of each particle. The particle size and the number of elongated particles were then computed in an automated form. The laboratory test results were compared with the results obtained from a machine learning-based algorithm. The results so obtained on the segregation of particles in various sizes give significant accuracy with a deviation of ±1 number of particles. The method so developed can be used by researchers and field engineers for the determination of the physical parameters of aggregates.