基于深度学习的食用油灌装质量检测系统

介绍了一种利用深度学习算法进行食用油灌装质量检测的系统,基于深度学习有监督物体识别网络对食用油生产线进行从原料至销售的全流程包装缺陷检测,具体功能包括瓶口缺陷检测、瓶盖缺陷检测、瓶身喷码缺陷检测、贴标缺陷检测、装箱点数检测。相比于传统机器视觉检测方案,该系统具有无需做图像预处理、检测精度高、参数设置简单、算法泛化能力强、开发周期短的优点,可实现食用油生产包装质量检测的全面自动化。     

基于深度学习的织物疵点检测研究进展

为提高疵点检测的准确性和通用性,实现使用简洁而有效的形式对织物图像的特点和疵点的本质特征进行综合表达,首先,介绍了深度学习技术,对引入了深度学习的疵点检测方法进行综述,同时对深度学习与疵点检测的内在关系进行阐述;然后,分析总结了深度学习的概念及代表性的计算模型,并对引入深度学习的疵点检测方法进行归纳、总结和分类;最后,对典型的方法进行了分析,讨论了各种方法的优缺点,并对未来的研究趋势进行了展望。指出：随着深度学习的发展,探索更加通用的检测方法是推进深度学习在织物疵点检测领域应用的努力方向。     

基于机器视觉+深度学习目标检测的皮革表面缺陷检测系统研究

针对皮革表面缺陷人工检测效率低、准确率低等问题,基于机器视觉和深度学习算法等构建了一种皮革表面缺陷检测系统.对该系统的主要框架及核心功能进行分析,以一般皮革表面光学检测系统为对象进行检测精度与检测效率对比.结果表明,基于机器视觉+深度学习的皮革表面缺陷检测系统检测精度更高,在应用初期的检测效率与一般检测系统较为接近,但随着应用时间的增长,系统检测效率优势也会逐渐显现.     

基于深度学习的进出口水产品品种鉴别技术研究进展

在经济全球化的大背景下,我国进出口贸易发展迅速,其中水产品占据很大的份额.随着进出口水产品的品种越来越丰富,这在一定程度上增加了水产品品种鉴别的难度,比如在进出口贸易过程中,部分企业受利益驱使,存在以假充真、以次充好等问题,严重影响了后续进出口贸易的发展.深度学习作为一种新型神经网络技术,可以充分模仿人脑机制的数据进行...     

基于深度学习与机器视觉的起重机吊装安全监测方法

随着我国经济的快速发展,各类大型工程层出不穷,对起重机吊装作业的需求不断增加。然而,吊装作业过程中依然存在众多的安全隐患,极易造成人员伤亡等安全事故。因此,该文提出一种基于深度学习和机器视觉的起重机吊装安全监测方法。将深度学习与机器视觉相结合对监控图像中的被吊物和工人进行识别和定位,同时可自主判断工人是否佩戴安全帽。根据监测模型的识别和定位信息,获得工人与被吊物之间的空间关系,为起重机吊装过程提供安全预警信息。为了提高所提方法的实用性和便携性,开发一个起重机吊装安全智能监测系统,不仅可以实时显示监控图像的识别结果,而且能够输出场景的语义描述、发出安全预警信号。     

基于深度学习的智能焊接技术研究

随着我国信息技术的普及,工业生产领域迎来了黄金发展时期,智能焊接技术在其中的应用频率处于上升趋势,但是仍然存在一些固有问题限制其发展,比如生产环境恶劣、系统标度大、双目系统特征匹配精度低等。在此背景下,基于深度学习的智能焊接技术应运而生。基于深度学习的智能焊接技术由三部分内容构成,分别是焊缝跟踪技术、双目立体视觉匹配技术以及目标识别技术,文章对此进行了详细论述,以提高工业生产效率与质量。     

基于深度相机的障碍物识别

为实现机器视觉系统的障碍物信息进行探测和识别,文章提出利用深度相机识别障碍物的方法:以障碍物和平面的距离变化梯度不同为主要判别方式,搭配孔洞填充,中值滤波,漫水填充,数学形态学运算,帧间运算等噪声抑制算法,实现对障碍物的探测。根据周围环境信息提取具有一定形状特征的典型目标物。实验结果表明,此方法能有效分别障碍物和平面。     

蜂蜜品种鉴别方法的研究进展

本研究利用氨基酸分析仪对我国3个不同品种单花蜜洋槐蜜、椴树蜜和油菜蜜,共计110个蜂蜜样品中17种氨基酸含量进行分析,并基于氨基酸含量结合统计分析方法进行蜂蜜种类鉴别分析。结果表明,油菜蜜中水解氨基酸含量高于椴树蜜和洋槐蜜,16种氨基酸含量在三个蜂蜜品种间存在差异。主成分分析（principle component analysis,PCA）结果表明不同植物源蜂蜜具有聚类趋势,偏最小二乘法判别分析（Partial least squares discriminant analysis,PLS-DA）结果表明油菜蜜可以和洋槐蜜和椴树蜜区分开来。线性判别分析（Linear discriminant analysis,LDA）结果表明3种蜂蜜整体判别率为92.7%,油菜蜜的判别率为92.3%。本研究为油菜蜜、洋槐蜜和椴树蜜分类鉴别提供数据支撑和参考依据。     

基于深度学习的条烟分拣线上烟包错配识别系统的构建

【背景】条烟分拣线上,条烟长边相邻并排摆放形成一层,多层叠加形成烟包,与订单相比,烟包可能存在少烟、多烟、品规错误等问题,目前采用的人工检查方式效率较低,且难以完全避免错误发生。本研究的目的是构建烟包错配识别系统。【方法】采用由工业相机镜头和光源构成的机器视觉系统采集成品烟包侧面与顶面图像,以基于深度学习的物体定位和识别技术获取烟包中条烟的数量与品规,与物流上位系统订单数据比对,自动识别与提示错误烟包。【结果】（1）实际使用中烟包识别成功率≥99.99%,识别耗时≤300ms。识别过程与原有工作步骤并行,增加识别系统不降低分拣效率。（2）系统上线运行至今有效避免了烟包连续出错和返工问题。（3）识别系统可以减轻搬运工人的工作负担,进而提高工作效率。【结论】采用深度学习机器视觉系统自动化识别烟包品规,可以提升烟草物流条烟分拣的质量和效率。     

Computer-vision classification of corn seed varieties using deep convolutional neural network

Automated classification of seed varieties is of paramount importance for seed producers to maintain the purity of a variety and crop yield. Traditional approaches based on computer vision and simple feature extraction could not guarantee high accuracy classification. This paper presents a new approach using a deep convolutional neural network (CNN) as a generic feature extractor. The extracted features were classified with artificial neural network (ANN), cubic support vector machine (SVM), quadratic SVM, weighted k-nearest-neighbor (kNN), boosted tree, bagged tree, and linear discriminant analysis (LDA). Models trained with CNN-extracted features demonstrated better classification accuracy of corn seed varieties than models based on only simple features. The CNN-ANN classifier showed the best performance, classifying 2250 test instances in 26.8 s with classification accuracy 98.1%, precision 98.2%, recall 98.1%, and F1-score 98.1%. This study demonstrates that the CNN-ANN classifier is an efficient tool for the intelligent classification of different corn seed varieties.     

Knitted fabric structure recognition based on deep learning

In knitted fabric structure recognition, the recognition rate is influenced by uneven light, fabric hairiness, fabric rotation, fabric thickness variation, yarn deviation, and loop deformation. To solve this problem, a method for recognizing knitted fabric structure based on deep learning is proposed. Firstly, sample images of fabrics are captured and a knitted fabric structure image database is established. Secondly, based on deep convolution neural network and transfer learning, the bvlc_reference_caffenet model trained by AlexNet is used as the pre-trained network. Then the pre-trained parameters of the network are transferred to the target data-set and the network is trained. Finally, the knitted fabric structure is recognized by the trained network. Experiment results show that the proposed recognition method is robust, which can overcome the influence of fabric rotation, fabric hairiness and uneven light, and achieves a high recognition rate.     

基于深度学习的牛肉大理石纹智能分级研究

目的开发客观、准确、无损的基于深度学习的牛肉大理石纹智能化分级技术。方法将深度学习的图像识别方法应用于牛肉大理石纹的特征提取和分类上,并进行相应的调试和学习。结果通过计算机调试和学习,评级正确率分别达到84.2%(一级)、89.4%(二级)、81.9%(三级)、84.1%(四级)、82.6%(五级)。各级牛肉的识别率均在80%以上,识别时间都在1 s以内,达到了预期目标。结论将深度学习的图像识别方法应用于牛肉大理石纹的特征提取和分类上,评级准确率非常高,且随着图片数据库样本数的不断增多,其识别的准确度将不断提高,可进行大量推广使用。     

基于机器视觉技术预测玉米质量等级的方法研究

国家标准规定玉米的质量定等指标为容重,为了实现利用机器视觉快速预测玉米等级,采用自行构建的基于机器视觉技术的玉米检测系统获取4种不同等级的玉米籽粒图像,通过均值滤波、最大类间方差法和形态学运算对玉米籽粒和背景进行处理、分割和特征参数的选取,并采用主成分分析法确定图像特征信息的最佳主成分因子数,建立以玉米容重为基础的8-21-4三层BP神经网络质量等级识别模型。结果表明:利用BP神经网络对基于完整籽粒图像和籽粒横切面图像的玉米等级的总体识别率均在90%以上,因此利用该模型对玉米等级的检测识别具有较高的可行性。     

基于玉米籽粒近红外光谱的品种与产地识别研究

为考察近红外光谱对玉米种子的品种识别与产地识别性能,采集了8个玉米品种波长范围为12 000~4 000 cm-1的近红外光谱数据,并基于此数据研究了基于PCA的光谱数据特征的提取方法,并探讨了神经网络(ANN)和支持向量机模型(SVM)在品种识别上的性能,进一步研究了玉米品种的产地识别技术,且比较了传统可见光图像的品种识别。研究发现:基于近红外的玉米品种识别,在6个主分量的情况下整体上性能达到90%以上;SVM算法较ANN算法稳定可靠,更适合于小样本情况下的光谱分析;基于光谱的品种识别与基于可见光图像的品种识别效果相当;另外发现同一品种在不同产地上其光谱特征差别较大,据此可以应用光谱进行产地鉴别,鉴别力达到95%以上。本研究所构建的方法对玉米品种识别和产地识别具有积极意义。     

基于深度学习的布匹疵点检测方法

为快速、准确检测布匹疵点,提出以深度学习目标检测框架YOLOv4为基础的布匹疵点检测方式,首先将5种常见疵点图像(吊经、百脚、结点、破洞、污渍)进行预处理,然后将图像输入到YOLOv4算法中进行分类。YOLOv4采用CSPDarknet53作为主干网络提取疵点特征,SPP模块、FPN+PAN的方式作为Neck层进行深层疵点特征提取,预测层采用3种尺度预测方式,对不同大小的疵点进行检测。研究结果表明:经600个测试集样本的验证,该方法对疵点图像的检测准确率达95%,检测单张疵点图像的速率为33 ms。与SSD、Faster R-CNN、YOLOv3方法进行比较,采用YOLOv4方法准确率更高,速度更快。     

基于深度学习的发霉花生识别技术

目的:快速、无损地识别发霉花生,提高发霉花生的识别效率。方法:采用光谱仪采集高光谱花生数据,利用深度学习技术识别霉变花生,建立Hypernet PRMF模型,并以Deeplab v3+、Segnet、Unet和Hypernet作为对照模型进行比较。将所提出的花生识别指数融合到高光谱图像中,作为数据特征预提取。同时将构建的多特征融合块集成到控制模型中以提高发霉花生识别效率。结果:所有模型的平均像素精度均超过了87%。Hypernet-PRMF模型的检测精度最高,达到90.35%,同时对于整个花生数据集,Hypernet-PRMF的错误识别率较低,可以有效识别图中所有的发霉花生。结论:基于深度学习所建立的Hypernet-PRMF模型具有较高的像素精度与检测精度,可有效识别发霉花生。     

Identifying Potato Varieties Using Machine Vision and Artificial Neural Networks

The objective of this study is to develop a method for identifying and discriminating 10 potato varieties by combining machine vision and artificial neural network methods. The potato varieties include Agria, Savalan, Florida, Fontaneh, Natasha, Verona, Karso, Elody, Satina, and Emrad. A total number of 72 characteristic parameters specifying color, textural, and morphological features are found among these varieties. By using principal component analysis, 16 principal features are selected for identifying and discriminating potato varieties. The data obtained from image processing were classified using linear discriminant analysis and non-linear artificial neural network method. The accuracy of discriminant analysis were 73.3, 93.3, 73.3, 40, 73.3, 73.3, 66.7, 80, 40, and 53.3%, respectively, for the varieties used in this study. The classification accuracy was improved by 100% for all the varieties using neural network analysis and the correct classification ratio was 100% using this method. It is revealed from the results that machine vision technique and neural network analysis could identify potato varieties with acceptable accuracy.     

Monitoring the respiratory behavior of multiple cows based on computer vision and deep learning

Automatic respiration monitoring of dairy cows in modern farming not only helps to reduce manual labor but also increases the automation of health assessment. It is common for cows to congregate on farms, which poses a challenge for manual observation of cow status because they physically occlude each other. In this study, we propose a method that can monitor the respiratory behavior of multiple cows. Initially, 4,000 manually labeled images were used to fine-tune the YOLACT (You Only Look At CoefficienTs) model for recognition and segmentation of multiple cows. Respiratory behavior in the resting state could better reflect their health status. Then, the specific resting states (lying resting, standing resting) of different cows were identified by fusing the convolutional neural network and bidirectional long and short-term memory algorithms. Finally, the corresponding detection algorithms (lying and standing resting) were used for respiratory behavior monitoring. The test results of 60 videos containing different interference factors indicated that the accuracy of respiratory behavior monitoring of multiple cows in 54 videos was >90.00%, and that of 4 videos was 100.00%. The average accuracy of the proposed method was 93.56%, and the mean absolute error and root mean square error were 3.42 and 3.74, respectively. Furthermore, the effectiveness of the method was analyzed for simultaneous monitoring of respiratory behavior of multiple cows under movement, occlusion disturbance, and behavioral changes. It was feasible to monitor the respiratory behavior of multiple cows based on the proposed algorithm. This study could provide an a priori technical basis for respiratory behavior monitoring and automatic diagnosis of respiratory-related diseases of multiple dairy cows based on biomedical engineering technology. In addition, it may stimulate researchers to develop robots with health-sensing functions that are oriented toward precision livestock farming.