Impurity detection of juglans using deep learning and machine vision
Impurity detection of juglans using deep learning and machine vision
复制标题
使用深度学习和机器视觉检测核桃的杂质
DOI：
10.1016/j.compag.2020.105764
复制
发表时间：
2020-11-01
影响因子：
8.3
通讯作者：
Zhang, Yinsheng
中科院分区：
文献类型：
--
作者：
Rong, Dian;Wang, Haiyan;Zhang, Yinsheng
在众多食品安全控制和质量检测应用中，杂质检测是定量图像分析的关键。由于不同姿态的异物形状和颜色复杂，利用机器视觉快速检测核桃仁中的杂质仍然面临挑战。一些传统的检测方法需要专门设计约束条件和手动模型参数，检测性能差，模型维护成本高。近年来，由于基于深度学习的方法能够直接从训练数据中学习特征，深度学习已成为不同研究领域的研究热点。在这项研究中，我们首次提出了两级卷积网络来实时完成图像分割和Juglan图像中杂质的检测。本文提出的基于多尺度残差完全卷积网络的图像分割方法和基于卷积网络的分类方法能够自动分割图像，同时检测出不同大小的杂质(如树叶碎片、废纸、塑料碎片和金属零件)。该深度学习方法避免了人工提取特征，不仅克服了在线图像中核桃和异物的粘连现象，而且适应了白色传动带表面磨损的干扰，避免了实际工厂环境中的误检，因此更加简单有效。该方法能够正确分割测试图像中99.4%的目标区域，正确分类验证图像中96.5%的异物，正确检测100.0%的测试图像。每幅图像的分割和检测处理时间小于60ms。未来的工作将集中在利用多波成像进行深度学习和分选机械控制方面。
Impurity detection is crucial for quantitative image analysis in numerous food safety control and quality inspection applications. Rapid impurity detection of juglans using machine vision still faces challenge due to the complex shapes and color of foreign objects in different postures. Some traditional detection methods require expertly designed constraints and manual model parameters, and they have the poor detection performance and high model maintenance costs. In recent years, deep learning has become a focus in different research fields, because methods based on deep learning are able to directly learn features from training data. In this study, we originally proposed the two-stage convolutional networks to finish the image segmentation and detection of impurities in juglans images in real-time. The proposed segmentation method based on multiscale residual fully convolutional networks and classification method based on convolutional networks automatically can segment images and detect different-sized impurities (e.g., leaf debris, paper scraps, plastic scraps and metal parts) at the same time. The proposed deep-learning method is simpler and more effective, because it avoids extracting features manually, and it not only overcomes the conglomeration phenomenon between juglans and foreign objects in inline images, but also adapts to the disturb from surface abrasion damage on the white transmission belt to avoid error detection in the real factory environment. The proposed method is able to correctly segment 99.4% of the object regions in the test images and to correctly classify 96.5% of the foreign objects in the validation images and correctly detect 100.0% of test images. The segmentation and detection processing time of each image was less than 60 ms. Future work will focus on deep learning using multi-wave imaging and the sorting mechanical control.