Crushed aggregates play an important role in the safe and reliable operation of our nations railroad infrastructure. Over time, aggregates become fouled by finer particles, and if not monitored and remedied, this fouling can lead to catastrophic accidents such as train derailments. A handheld inspection system that can accurately perform a quantitative measure of aggregate gradation and degree of fouling is the focus of this project. This system will produce an immediate measurement of aggregate fouling condition without the need for material sampling in the field and laboratory testing. We will harness the integrated camera, global positioning sensor, micro-processor and communication interfaces of modern mobile devices to complete image analysis locally for immediate assessment of conditions. A key element of this work will be an external sensor to augment computer vision detection and classification of aggregate particles. This external sensor leverages recent advancements in light detection and ranging hardware with novel optical design to eliminate manual and discrete scale references in acquired images. The image analysis algorithms are based on cutting edge work on aggregate particle detection and segmentation.In combination, the system overcomes limitations in current methods and may be extended to many other infrastructure inspection tasks.
