Hyperspectral Imaging for Identifying Foreign Objects on Pork Belly

2025-03-20

AI Summary

This study addresses the challenge of detecting surface contaminants—such as plastic and rubber—on pork belly in industrial production lines, where conventional computer vision methods fail due to low visual contrast. We propose a pixel-level semantic segmentation framework integrating near-infrared hyperspectral imaging (900–1700 nm) with a lightweight Vision Transformer (ViT). To our knowledge, this is the first application of a lightweight ViT to hyperspectral contaminant segmentation in meat processing. The method incorporates adaptive spectral preprocessing and a channel-wise attention mechanism to mitigate interference from spectral similarity, ambient temperature fluctuations, and system noise. Evaluated on real-world production-line data, the approach achieves an average intersection-over-union (mIoU) of >98.2% and operates at 23 frames per second (FPS), significantly improving both robustness and real-time performance. This work delivers a deployable AI solution for quality control in meat processing.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Web Mining and Content Analysis: Web data integration and cleaning

Abstract

Ensuring food safety and quality is critical in the food processing industry, where the detection of contaminants remains a persistent challenge. This study presents an automated solution for detecting foreign objects on pork belly meat using hyperspectral imaging (HSI). A hyperspectral camera was used to capture data across various bands in the near-infrared (NIR) spectrum (900-1700 nm), enabling accurate identification of contaminants that are often undetectable through traditional visual inspection methods. The proposed solution combines pre-processing techniques with a segmentation approach based on a lightweight Vision Transformer (ViT) to distinguish contaminants from meat, fat, and conveyor belt materials. The adopted strategy demonstrates high detection accuracy and training efficiency, while also addressing key industrial challenges such as inherent noise, temperature variations, and spectral similarity between contaminants and pork belly. Experimental results validate the effectiveness of hyperspectral imaging in enhancing food safety, highlighting its potential for broad real-time applications in automated quality control processes.

Problem

• Detects foreign objects on pork belly using hyperspectral imaging.
• Improves food safety by identifying undetectable contaminants.
• Enhances automated quality control with high accuracy and efficiency.

Innovation

• Hyperspectral imaging for contaminant detection
• Lightweight Vision Transformer for segmentation
• Near-infrared spectrum data analysis