Size and Shape Determination of Riprap and Large-sized Aggregates Using Field Imaging

Huang, Haohang; Luo, Jiayi; Tutumluer, Erol; Hart, John M.; Qamhia, Issam

Permalink

https://hdl.handle.net/2142/106037 Copy

Description

Title

Size and Shape Determination of Riprap and Large-sized Aggregates Using Field Imaging

Author(s)

• Huang, Haohang
• Luo, Jiayi
• Tutumluer, Erol
• Hart, John M.
• Qamhia, Issam

Issue Date

2020-01

Keyword(s)

• Riprap
• Large-sized Aggregates
• Field Imaging
• Computer Vision
• Deep Learning
• Object Detection
• Image Segmentation
• Size and Morphological/Shape Analyses
• Stockpile Aggregate Imaging

Abstract

Riprap rock and large-sized aggregates are extensively used in transportation, geotechnical, and hydraulic engineering applications. Traditional methods for assessing riprap categories based on particle weight may involve subjective visual inspection and time-consuming manual measurements. Aggregate imaging and segmentation techniques can efficiently characterize riprap particles for their size and morphological/shape properties to estimate particle weights. Particle size and morphological/shape characterization ensure the reliable and sustainable use of all aggregate skeleton materials at quarry production lines and construction sites. Aggregate imaging systems developed to date for size and shape characterization, however, have primarily focused on measurement of separated or non-overlapping aggregate particles. This research study presents an innovative approach for automated segmentation and morphological analyses of stockpile aggregate images based on deep-learning techniques. As a project outcome, a portable, deployable, and affordable field-imaging system is envisioned to estimate volumes of individual riprap rocks for field evaluation. A state-of-the-art object detection and segmentation framework is used to train an image-segmentation kernel from manually labeled 2D riprap images in order to facilitate automatic and user-independent segmentation of stockpile aggregate images. The segmentation results show good agreement with ground-truth validation, which entailed comparing the manual labeling to the automatically segmented images. A significant improvement to the efficiency of size and morphological analyses conducted on densely stacked and overlapping particle images is achieved. The algorithms are integrated into a software application with a user-friendly Graphical User Interface (GUI) for ease of operation. Based on the findings of this study, this stockpile aggregate image analysis program promises to become an efficient and innovative application for field-scale and in-place evaluations of aggregate materials. The innovative imaging-based system is envisioned to provide convenient, reliable, and sustainable solutions for the on-site quality assurance/quality control (QA/QC) tasks related to riprap rock and large-sized aggregate material characterization and classification.

Publisher

Illinois Center for Transportation/Illinois Department of Transportation

Series/Report Name or Number

FHWA-ICT-20-002

ISSN

0197-9191

Type of Resource

text

Language

en

Permalink

http://hdl.handle.net/2142/106037

DOI

10.36501/0197-9191/20-003

Sponsor(s)/Grant Number(s)

IDOT-R27-182

Copyright and License Information

No restrictions. This document is available through the National Technical Information Service, Springfield, VA 22161.

Owning Collections