Fractal texture analysis of bread crumb digital images

A fractal texture analysis technique was applied to bread crumb digital images. Fractal dimensions obtained from several methods (fractional Brownian motion, frequency domain, relative differential box-counting, morphological fractal, mass fractal and random walks methods) were investigated in order to determine their capability to accurately describe the surface roughness of bread crumb images or the visual appearance of bread crumb in meaningful terms. A total of 500 bread crumb images of different porosity and grain quality were analysed. It was found that bread crumb appearance could be effectively quantified by the fractal dimension of its digital image. Correlations of fractal dimensions with mean cell area, standard deviation of cell area and void fraction were variable for the fractal methods. While the mass fractal method measured better crumb heterogeneity, other methods quantified coarseness, cell–cell wall ruggedness and cell wall tortuosity. A vector comprising fractal dimensions would objectively depict crumb grain and would allow comparisons between different bread crumb images.     

Investigation of image registration method for the multi-directional image fusion of woven fabrics

Abstract

Traditional single-side scanning or single-vision image acquisition methods have the limitation of incomplete information caused by the existence of blind spots. To collect the complete texture information of fabric images, a new multi-vision image acquisition and the related fusion method is developed to solve this problem. However, linear addition of image sequences acquired from multiple directions cannot achieve a good result of image fusion, it is necessary to conduct the image fusion based on the image registration between images at pixel level. Therefore, a new multi-directional digital image acquisition system for woven fabrics is established in this article; one set of image fusion algorithm based on image registration is proposed for the image enhancement of fabric. Fabric texture images are digitized by means of multi-directional vision imaging instead of unidirectional imaging, the structural information of fabric texture could be enhanced using image registration and fusion technology and the indexing and localization of texture corresponding points could be controlled using matching points or control points. Our experimental results show that the proposed method could be used to merge the effective information from the multi-directional vision images completely, it has the potential application for the rendering of woven fabrics using image driven virtual reality enhancement.     

Assessing optimal frequency for image acquisition in computer vision systems developed to monitor feeding behavior of group-housed Holstein heifers

Computer vision systems have emerged as a potential tool to monitor the behavior of livestock animals. Such high-throughput systems can generate massive redundant data sets for training and inference, which can lead to higher computational and economic costs. The objectives of this study were (1) to develop a computer vision system to individually monitor detailed feeding behaviors of group-housed dairy heifers, and (2) to determine the optimal frequency of image acquisition to perform inference with minimal effect on feeding behavior prediction quality. Eight Holstein heifers (96 ± 6 d old) were housed in a group and a total of 25,214 images (1 image every second) were acquired using 1 RGB camera. A total of 2,209 images were selected and each animal in the image was labeled with its respective identification (1–8). The label was annotated only on animals that were at the feed bunk (head through the feed rail). From the labeled images, 1,392 were randomly selected to train a deep learning algorithm for object detection with YOLOv3 (“You Only Look Once” version 3) and 154 images were used for validation. An independent data set (testing set = 663 out of the 2,209 images) was used to test the algorithm. The average accuracy for identifying individual animals in the testing set was 96.0%, and for each individual heifer from 1 to 8 the accuracy was 99.2, 99.6, 99.2, 99.6, 99.6, 99.2, 99.4, and 99.6%, respectively. After identifying the animals at the feed bunk, we computed the following feeding behavior parameters: number of visits (NV), mean visit duration (MVD), mean interval between visits (MIBV), and feeding time (FT) for each heifer using a data set composed by 8,883 sequential images (1 image every second) from 4 time points. The coefficient of determination (R²) was 0.39, 0.78, 0.48, and 0.99, and the root mean square error (RMSE) were 12.3 (count), 0.78, 0.63, and 0.31 min for NV, MVD, MIBV, and FT, respectively, considering 1 image every second. When we moved from 1 image per second to 1 image every 5 (MIBV) or 10 (NV, MDV, and FT) s, the R² observed were 0.55 (NV), 0.74 (MVD), 0.70 (MIBV), and 0.99 (FT); and the RMSE were 2.27 (NV, count), 0.38 min (MVD), 0.22 min (MIBV), and 0.44 min (FT). Our results indicate that computer vision systems can be used to individually identify group-housed Holstein heifers (overall accuracy = 99.4%). Based on individual identification, feeding behavior such as MVD, MIBV, and FT can be monitored with reasonable accuracy and precision. Regardless of the frequency for optimal image acquisition, our results suggested that longer time intervals of image acquisition would reduce data collecting and model inference while maintaining adequate predictive performance. However, we did not find an optimal time interval for all feeding behavior; instead, the optimal frequency of image acquisition is phenotype-specific. Overall, the best R² and RMSE for NV, MDV, and FT were achieved using 1 image every 10 s, and for MIBV it was achieved using 1 image every 5 s, and in both cases model inference and data storage could be drastically reduced.     

棉纤维长度测量技术的创新性探讨

将棉纤维试样用包埋技术进行介质固定,然后进行横向连续10μm渐进切割,每切割一次用CCD图像采集系统摄取横切面放大图像,并输入计算机。通过图像软件技术对数据进行处理,实现单根棉纤维空间立体再造。然后依据空间立体图像,通过计算而得到每根纤维的真实拉直长度。依据新的长度测量技术,可统计计算所有试样中棉纤维的详细长度分布,是纤维长度测量理论与测量技术的一次创新,基于此理论与技术开发的短纤维率测量仪将填补国内空白,纺织应用前景广阔。     

黑小麦粉的流变学特性及在食品中的应用研究

研究了黑小麦粉的流变学特性及在面条和面包中的应用。研究结果表明,黑小麦粉属于中等筋力的面粉,可适量地应用于面条和面包的生产中。     

红枣输送运动及其图像采集的可靠性分析

滚子式输送机构在机器视觉分级系统中应用最为广泛。通过对滚子输送机构中的红枣进行受力分析和运动分析,获取了红枣从刚接触摩擦带由静止加速到与滚子接触点处线速度相同时水平前进距离的数学模型;获取了红枣加速转动过程结束后,定距离间隔采集红枣图像时红枣转动角度的数学模型。以新疆骏枣为例,通过确定合适的传输速度、相机安装位置和相机对枣的图像采集次数,能够采集到红枣整个表面信息,从而保证图像采集的可靠性。     

基于卷积神经网络模型的裂纹鸡蛋图像识别

目的:提高基于计算机视觉的鸡蛋裂纹检测方法的准确性和运行效率。方法:使用禽蛋模拟撞击设备得到裂纹鸡蛋,并通过鸡蛋动态图像采集设备采集不同角度裂纹鸡蛋和完好鸡蛋图像,然后以原始图像和经预处理后图像分别建立用于裂纹鸡蛋图像识别的YOLO-v5、ResNet和SuffleNet模型,并比较不同模型识别准确度以及对未经预处理图像的适应性。结果:YOLO-v5、ResNet和SuffleNet模型均可有效识别经过预处理的裂纹鸡蛋图像,其验证集准确率分别为98.8%,97.8%,99.4%。对于未经预处理的裂纹鸡蛋,ResNet模型判别准确率较低,而SuffleNet模型对其适应性较好,判别准确度超过99%。结论:在卷积神经网络模型中,SuffleNet模型适用于裂纹鸡蛋图像的识别,且采集的图像无需进行预处理。     

一种测定羊绒颜色纯度的计算机视觉系统

传统测定羊绒纯度的手工方法受主观因素影响,精度和速度不能满足生产需要.文章提出了一种由纤维分匀器、图像获取系统和图像处理计算机组成的、能客观测定羊绒颜色纯度的计算机视觉系统.由实验结果的对比可以看出本系统的研究成果是可行的.文章还对系统采用的检测样品的预处理、羊绒中杂色毛的识别与评定算法等关键技术进行了重点介绍.     

Increasing the health benefits of bread: Assessment of the physical and sensory qualities of bread formulated using a renin inhibitory Palmaria palmata protein hydrolysate

This study examined the textural and sensory effects of bread formulated using a Palmaria palmata protein hydrolysate with in vitro renin inhibitory properties. Four formulations were examined; a wheat flour control bread, a control bread containing 4% P. palmata protein hydrolysate, a buckwheat bread consisting of a blend of 70% wheat: 30% buckwheat, and a fourth formulation of the buckwheat bread containing 4% P. palmata hydrolysate. The volume, colour, texture profile, moisture, crumb structure, sensory attributes and renin inhibitory activity of all four baked breads are presented. Bright field light microscopy and confocal scanning laser microscopy images using a new triple labelling protocol were also obtained to visualise the main ingredients of the breads. It was observed that the addition of 4% P. palmata protein hydrolysate to the wheat bread control did not affect the texture or sensory properties of the bread to a large degree. Furthermore, when compared to the control, the bread containing the hydrolysate retained renin inhibitory bioactivity after the baking process. This study highlights that baked products may be a suitable delivery vehicle for bioactive compounds. The study demonstrates that the health value of bread was increased through the addition of a seaweed protein hydrolysate.     

Characterization of bread breakdown during mastication by image texture analysis

The methods currently used to characterise food breakdown during mastication are not easily applicable to cohesive and heterogeneous products such as white bread (baguettes). During this study, we investigated the applicability of image texture analysis to characterising the kinetics of bread bolus formation during chewing. Food boluses were collected from five subjects chewing four different breads after 10, 20 and 30 chewing cycles or at swallowing. Images were acquired and analysed using the Grey level co-occurrence matrix (GLCM) method. Food boluses were successfully discriminated: 60–73% of the images were correctly classified into their respective chewing cycles. Incorrect classifications arose from overlapping between subsequent cycles. Among the texture features the contrast was identified as being the best marker of food degradation. Different kinetics of bolus formation were observed between breads. This method revealed specific patterns of food degradation between subjects, which could be explained by their mastication behaviour and chewing efficiency.     

一种基于聚类特征的Faster R-CNN粮仓害虫检测方法

本文基于Faster R-CNN模型提出复杂背景下粮仓害虫的检测识别方法。将六种常见的储粮害虫(豆象、谷蠹、米象、锯谷盗、赤拟谷盗、锈赤扁谷盗)分别以大米、小米为背景,建立了真实背景下粮仓害虫图像数据集SGI-6。SGI-6中包括网络获取图像、显微镜采集图像和单反拍摄图像三种多目标尺度的数据集。根据粮仓害虫的小目标特性,使用聚类算法改进Faster R-CNN模型的区域提案网络,来提取这些图像中含有害虫的区域,并对这些区域中的害虫进行分类。实验结果表明,该方法能够在储粮条件下检测和识别粮仓害虫,且其平均准确率(mAP)达到96.63%。     

Determination of a representative volume element based on the variability of mechanical properties with sample size in bread

Quantitative analysis of food structure is commonly obtained by image analysis of a small portion of the material that may not be the representative of the whole sample. In order to quantify structural parameters (air cells) of 2 types of bread (bread and bagel) the concept of representative volume element (RVE) was employed. The RVE for bread, bagel, and gelatin-gel (used as control) was obtained from the relationship between sample size and the coefficient of variation, calculated from the apparent Young's modulus measured on 25 replicates. The RVE was obtained when the coefficient of variation for different sample sizes converged to a constant value. In the 2 types of bread tested, the tendency of the coefficient of variation was to decrease as the sample size increased, while in the homogeneous gelatin-gel, it remained always constant around 2.3% to 2.4%. The RVE resulted to be cubes with sides of 45 mm for bread, 20 mm for bagels, and 10 mm for gelatin-gel (smallest sample tested). The quantitative image analysis as well as visual observation demonstrated that bread presented the largest dispersion of air-cell sizes. Moreover, both the ratio of maximum air-cell area/image area and maximum air-cell height/image height were greater for bread (values of 0.05 and 0.30, respectively) than for bagels (0.03 and 0.20, respectively). Therefore, the size and the size variation of air cells present in the structure determined the size of the RVE. It was concluded that RVE is highly dependent on the heterogeneity of the structure of the types of baked products.     

基于太赫兹反射成像技术的花生空果无损检测

目的 基于太赫兹反射成像技术(Terahertz Reflection Imaging Method, THz-RIM)实现花生空果无损检测,为太赫兹技术在农产品及食品快速无损检测中的应用奠定基础。方法 利用太赫兹脉冲光谱仪TeraPulse 4000及反射成像附件和采集软件,在0.2 mm采集步长下对去壳花生仁进行太赫兹反射成像预实验,确定出正式实验的图像颜色模式为hot模式。人为制备若干个花育36号满仁花生果样本和空果样本(仅有一粒花生仁),在与预试验相同的条件下采集样本光谱图像,以时域/频域2点外相加成像(Integrated Time Slice)重构方式进行图像重构, 突出花生空果样本图像的显著几何特征。结果 通过太赫兹反射图像可以实现满仁与花生空果样本的检测区分,但图像细节仍然比较模糊,需进一步提高图像清晰度。采用数字图像方法进行图像增强,利用OpenCV-Python提高图像的亮度、色度、对比度及锐度。提升对比度后的图像可观测度最高,且空果图像中黑色部分面积占比最大,与含仁部分对比最为明显,可以实现空果及满仁花生果的快速区分。结论 探索结果表明,THz波对花生外壳具有良好的穿透性,利用THz-RIM技术结合OpenCV-Python数字图像处理方法进行花生空果无损检测具有可行性,利用太赫兹光谱技术对油料作物内部品质进行无损检测研究具有极大潜力。     

Suitability of various australian wheats for chinese-style steamed bread

Three methods of preparation of steamed bread on a laboratory scale were evaluated to determine the one most suitable for discriminating between flour samples. The preferred method was then applied to a number of wheat varieties at different protein levels to select the most suitable ones for this product, and to determine the key analytical parameters of the most suitable flour. Flour colour and protein content appeared to be more important than other physical and chemical properties. Flour samples with high protein content (greater than 12%) and strong doughs gave bread with a wrinkled surface and dark colour, whereas low protein (less than 10%) soft wheat flours gave poor texture and eating quality, although the surface was smooth. It is suggested that flours of medium protein content (10–12%) and medium dough strength are the most suitable for steamed bread. Processing conditions affected steamed bread quality. Blending of high protein hard strong wheats with soft weak wheat could be practical for preparing flour for steamed bread.     

应用Matlab 6.5软件对奶酪和面包样品的核磁共振成像进行T2分布重建的研究

本文论述了采用低场核磁共振成像技术（magnetic resonance imaging,MRI）中的spin3D成像序列,采用不同Tau值对奶酪和面包样品进行成像实验.应用Matlab软件编程对图像进行数据处理,并对T2分布进行重建,最后得到T2分布状态图像.为核磁共振技术在食品中的应用打下良好的基础.     

织物静态悬垂性测试过程中图像采集问题的研究

分别采用织物动态悬垂性风格仪和基于织物静态悬垂性风格仪自行组装的DIY仪器,对织物的静态悬垂性进行测试,并对测试过程中图像采集方式的选择和俯视死角问题进行分析。研究结果表明:当纯色织物的颜色与背景颜色反差较大或花色织物的图案花色与背景颜色相差较大时,宜采用上打光直接采集图像;当测试织物为半透明织物时,宜采用下打光直接采集图像;上打光顺光采集图像过程中出现俯视死角导致的图像采集不清晰问题,可通过增大支持盘半径或增大相机与试样间距的方法解决。     

含不同粗粮粉面包的营养、质构特性、风味化合物

以小麦粉为主要原料,分别添加一定量马铃薯全粉、荞麦粉、薏苡粉、燕麦粉制作粗粮面包,以氨基酸组成、质构、色泽、风味为评价指标,分析粗粮面包营养特性(主要是游离氨基酸含量)、质构特性、色泽变化和挥发性风味化合物的特征,同时以普通面粉面包为对照。结果表明:从质构特性来看,粗粮粉添加量不应超过12%;除薏苡组外,其余组的氨基酸综合评分均高于对照组;粗粮组面包风味较对照组浓郁,风味物质的总量和种类均高于对照组;添加粗粮粉制作的面包会产生一些特别的风味物质,如壬醇、3-甲基十五烷、十八烷、植烷、3-甲基丁酸、2,6-二甲基吡嗪等。因此,添加适量的粗粮粉既能丰富面包的营养,又能产生浓郁的风味物质,而对面包色泽影响不大。     

纱线条干均匀性序列图像测量方法

为了更精确、连续地评价纱线的外观条干均匀性,提出一种基于序列图像的纱线条干均匀性测量方法。通过图像采集系统获取连续的纱线序列图像;根据模糊C-均值（FCM）聚类算法将图像进行阈值分割,得到纱线条干的二值图像;设定阈值去除图像中的孤立点、毛刺点,获得清晰的纱线条干图像,并计算图像中每行纱线直径。 为验证方法的准确性,对７ 种线密度的紧密纺纯棉纱进行了条干测试,并将结果与USTER® Tester 5-S800 条干测试仪的测试结果进行了对比。结果表明：序列图像方法测得的纱线条干均匀性与USTER 条干仪的测试结果高度正相关,证明本文的方法是准确可行的。