用于 AFB1 检测的多层空间-光谱分割网络

摘要：黄曲霉毒素是一种由黄曲霉产生的毒素，常见于花生中，对人体健康构成重大威胁。在本文中，我们提出了一个 可调整的Mul-U-Net模型，能够并行学习光谱和空间信息，通过调整中央空间U单元的数量来促进黄曲霉毒素的快速检 测。首先，通过将高光谱图像立方体输入到六个语义分割模型：Mul-U-Net，U-Net，FCN，SegNet，CNN和RefineNet中， 对AFB 进行语义分割。实验结果表明，Mul-U-Net优于其他模型，实现了95.64%的总体准确率和0.2079的RMSE。我 1 们进一步研究了不同空间 U 单元对模型性能的影响，发现最佳空间 U 计数与标签数量相匹配。在黄曲霉毒素数据集中， 穆尔-2 U-Net模型达到了96.12%的最高准确率。对Indian Pines和Pavia University数据集的进一步验证表明，Mul-16 U- Net和Mul-9 U-Net分别达到了98.12%和98.01%的最高准确率。本研究提出了一种鲁棒、高性能、可调的Mul-U-Net语 义分割网络，为花生中黄曲霉毒素的识别提供了一种有价值的新方法。

关键词：高光谱成像，语义分割，空间信息，光谱信息。

期刊：Journal of Food Composition and Analysis

DOI：https://doi.org/10.1016/j.jfca.2024.106790

设备：GaiaField

Multilayer spatial-spectral segmentation network for detecting AFB 1

Abstract：Aflatoxin, a toxin produced by Aspergillus flavus, is commonly found in peanuts and poses a significant threat to human health. In this paper, we propose an adjustable Mul-U-Net model, capable of learning both spectral and spatial information in parallel, facilitating rapid detection of aflatoxins by adjusting the number of central spatial U units. First, AFB 1 was semantically segmented by inputting hyperspectral image cubes into six semantic seg mentation models: Mul-U-Net, U-Net, FCN, SegNet, CNN, and RefineNet. Experimental results demonstrate that Mul-U-Net outperformed the other models, achieving an overall accuracy of 95.64 % and an RMSE of 0.2079. We further examined the influence of different spatial U units on model performance and found that the optimal spatial U count matched the number of labels. In the aflatoxin dataset, the Mul-2 U-Net model achieved the highest accuracy of 96.12 %. Further validation on the Indian Pines and Pavia University datasets showed that Mul-16 U-Net and Mul-9 U-Net achieved the highest accuracies of 98.12 % and 98.01 %, respectively. This study presents a robust, high-performance, and tunable Mul-U-Net semantic segmentation network, offering a valuable new approach for the identification of aflatoxins in peanuts.

Keywords：Aflatoxin B , Hyperspectral imaging, Semantic segmentation, Spatial information, Spectral information. 1