# Deep Learning-Based Automated Image Segmentation for Concrete Petrographic Analysis

arXiv: https://arxiv.org/abs/2005.10434 · HTML: https://mlsanity.com/paper/2005.10434/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 May 2020.

## Built on by (most cited first, 6 of 6)

- [Deep-XFCT: Deep learning 3D-mineral liberation analysis with micro X-ray fluorescence and computed tomography](https://mlsanity.com/paper/2205.13102/) (2022)
- [Artificial Intelligence in Concrete Materials: A Scientometric View](https://mlsanity.com/paper/2209.09636/) (2022)
- [DINOv2 Rocks Geological Image Analysis: Classification, Segmentation, and Interpretability](https://mlsanity.com/paper/2407.18100/) (2024)
- [An Efficient Instance Segmentation Approach for Extracting Fission Gas Bubbles on U-10Zr Annular Fuel](https://mlsanity.com/paper/2302.12833/) (2023)
- [Image-based Deep Learning for the time-dependent prediction of fresh concrete properties](https://mlsanity.com/paper/2402.06611/) (2024)
- [Medical Image Analysis for Detection, Treatment and Planning of Disease using Artificial Intelligence Approaches](https://mlsanity.com/paper/2405.11295/) (2024)

## Abstract (arXiv)

The standard petrography test method for measuring air voids in concrete (ASTM C457) requires a meticulous and long examination of sample phase composition under a stereomicroscope. The high expertise and specialized equipment discourage this test for routine concrete quality control. Though the task can be alleviated with the aid of color-based image segmentation, additional surface color treatment is required. Recently, deep learning algorithms using convolutional neural networks (CNN) have achieved unprecedented segmentation performance on image testing benchmarks. In this study, we investigated the feasibility of using CNN to conduct concrete segmentation without the use of color treatment. The CNN demonstrated a strong potential to process a wide range of concretes, including those not involved in model training. The experimental results showed that CNN outperforms the color-based segmentation by a considerable margin, and has comparable accuracy to human experts. Furthermore, the segmentation time is reduced to mere seconds.