ParticleSAM: Small Particle Segmentation for Material Quality Monitoring in Recycling Processes

2025-08-05

AI Summary

To address the low efficiency of manual quality inspection for dense, fine-grained particles (e.g., crushed bricks, aggregates) in construction recycled materials—and the poor performance of existing vision segmentation models (e.g., SAM) on small, densely packed objects—this paper proposes a novel image segmentation method tailored for dense granular scenes. Our approach features three key contributions: (1) an enhanced SAM architecture specifically optimized for small-object segmentation; (2) the first multi-particle benchmark dataset, automatically synthesized and annotated from single-particle images; and (3) a physics-aware particle image synthesis strategy that improves visual fidelity and generalization. Extensive experiments demonstrate substantial improvements over the original SAM: qualitative analysis confirms superior boundary accuracy and robustness, while quantitative evaluation shows over 12% gain in mean Intersection-over-Union (mIoU). The method exhibits strong transferability to other granular material inspection tasks.

Technology Category

• Computer Vision: Segmentation
• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The construction industry represents a major sector in terms of resource consumption. Recycled construction material has high reuse potential, but quality monitoring of the aggregates is typically still performed with manual methods. Vision-based machine learning methods could offer a faster and more efficient solution to this problem, but existing segmentation methods are by design not directly applicable to images with hundreds of small particles. In this paper, we propose ParticleSAM, an adaptation of the segmentation foundation model to images with small and dense objects such as the ones often encountered in construction material particles. Moreover, we create a new dense multi-particle dataset simulated from isolated particle images with the assistance of an automated data generation and labeling pipeline. This dataset serves as a benchmark for visual material quality control automation while our segmentation approach has the potential to be valuable in application areas beyond construction where small-particle segmentation is needed. Our experimental results validate the advantages of our method by comparing to the original SAM method both in quantitative and qualitative experiments.

Problem

• Automating quality monitoring of recycled construction materials
• Adapting segmentation models for small dense particle images
• Creating benchmark dataset for visual material quality control

Innovation

• Adapts segmentation model for small dense objects
• Automated data generation and labeling pipeline
• Benchmark dataset for material quality control