A simple method for weight estimation of whole herring (Clupea harengus) using planar X-ray imaging

Fish weight is one of the key parameters used in commercial sorting of fresh fish. The capability of on-line and accurate estimation of the weight of each individual fish has a direct impact on profitability for the fish processors. Planar X-ray images of whole herring (n=16) were taken using a commercial medical X-ray apparatus. The integral intensity of all pixels in each image was calculated and correlated with the weight of the respective fish producing a linear relationship with a high correlation factor (R2=0.992) and a calculated mean prediction error of 2.5%. Planar X-ray imaging combined with simple image processing was successfully used to predict weight of whole herring. The technique has a potential to be used industrially for on-line weight estimation in sorting/grading systems.     

Design of an Optimum Computer Vision‐Based Automatic Abalone (Haliotis discus hannai) Grading Algorithm

An automatic abalone grading algorithm that estimates abalone weights on the basis of computer vision using 2D images is developed and tested. The algorithm overcomes the problems experienced by conventional abalone grading methods that utilize manual sorting and mechanical automatic grading. To design an optimal algorithm, a regression formula and R² value were investigated by performing a regression analysis for each of total length, body width, thickness, view area, and actual volume against abalone weights. The R² value between the actual volume and abalone weight was 0.999, showing a relatively high correlation. As a result, to easily estimate the actual volumes of abalones based on computer vision, the volumes were calculated under the assumption that abalone shapes are half‐oblate ellipsoids, and a regression formula was derived to estimate the volumes of abalones through linear regression analysis between the calculated and actual volumes. The final automatic abalone grading algorithm is designed using the abalone volume estimation regression formula derived from test results, and the actual volumes and abalone weights regression formula. In the range of abalones weighting from 16.51 to 128.01 g, the results of evaluation of the performance of algorithm via cross‐validation indicate root mean square and worst‐case prediction errors of are 2.8 and ±8 g, respectively.     

Egg Freshness on-Line Estimation Using Machine Vision and Dynamic Weighing

This paper proposed a non-destructive method for online estimation of egg freshness based on machine vision and dynamic weighing. The machine vision system and the dynamic weighing system were developed for the measurement of egg external physical characteristic and weight. Digital signal processing was employed to denoise, analyze, and extract the effective feature information from the images and vibration signals. Then, the method of multiple linear regressions was used to build model for Haugh unit prediction using the parameters of long axis, minor axis, and weight. The performance of the predictive model using six variables was achieved, with R (correlation coefficient) of 0.8653 and RMSEP (root mean square error of prediction) of 3.7454 in prediction set. The speed of the system reaches four eggs per second for every measurement. Good consistence confirmed that the proposed method has a significant potential application in online estimation of the egg freshness.     

Rapid Assessment of Tomato Ripeness Using Visible/Near-Infrared Spectroscopy and Machine Vision

A non-destructive assessment using visible/near-infrared spectroscopy and machine vision has been investigated for measuring tomato ripeness. Relationship between spectral wavelengths and green grayscale value was evaluated by application of chemometrics techniques based on partial least squares (PLS) regression. The tomatoes were divided randomly into two groups: 170 fruits for calibration and 71 for prediction. An accurate estimation, measured with a correlation coefficient of 0.992 and root mean square errors of prediction (RMSEP) of 9.92, was obtained when using the developed PLS model built with 550–750 nm spectral range. The accuracies of calibration and validation models based on data measured in this band were 90.93 and 90.05%. The prediction accuracy for 150 external independent samples was 90.67%. The results show that it is possible to realize detection standardization of tomato maturity based on only visible spectroscopy (550–750 nm) and machine vision technologies. This detection method does not depend on a visual grading or other maturity indices as a reference. It highlights the potential of the method to determine tomato ripeness and the optimum harvest time.     

Adaptive detection of weft‐knitted fabric defects based on machine vision system

This paper describes a machine vision system for the detection of weft‐knitted fabric defects based on an adaptive pulse‐coupled neural network (PCNN) and Ridgelet transform. In order to classify defects according to their different texture features, two methods are implemented: an improved PCNN method to segment the defects such as hole and dropped stitch from background image and a Ridgelet transform method based on wavelet analysis to identify the defect such as course mark. In implementing the PCNN model, necessary parameters of PCNN model such as linking coefficient, connection weight, and iteration number are automatically calculated in accordance with the spatial distance of neurons, mean, and variance value of whole image, and the cross‐entropy criterion. The function of Ridgelet transform is to identify the straight line marks and fit the regression equation for simulating the course mark in the image. The Ridgelet transform model can be simplified as the combination of Radon and wavelet transforms. The parameters of detected line are acquired by wavelet analysis in Fourier semicircle region. The experiment materials were several plain and interlocked weft‐knitted fabrics with hole, dropped stitch, and course mark defects. The fabric images were captured by an area‐scan camera with a resolution of 600 × 800 pixels, and signal processing was controlled by a digital signal processing multiprocessor on the inspection machine. The validation tests proved that the system performed well.     

Prediction of the weight of Alaskan pollock using image analysis

Determining the size and quality attributes of fish by machine vision is gaining acceptance and increasing use in the seafood industry. Objectivity, speed, and record keeping are advantages in using this method. The objective of this work was to develop the mathematical correlations to predict the weight of whole Alaskan Pollock (Theragra chalcogramma) based on its view area from a camera. One hundred and sixty whole Pollock were obtained fresh, within 2 d after catch from a Kodiak, Alaska, processing plant. The fish were first weighed, then placed in a light box equipped with a Nikon D200 digital camera. A reference square of known surface area was placed by the fish. The obtained image was analyzed to calculate the view area of each fish. The following equations were used to fit the view area (X) compared with weight (Y) data: linear, power, and 2nd-order polynomial. The power fit (Y = A · X(B)) gave the highest R(2) for the fit (0.99). The effect of fins and tail on the accuracy of the weight prediction using view area were evaluated. Removing fins and tails did not improve prediction accuracy. Machine vision can accurately predict the weight of whole Pollock. Practical Application: The weight of Alaskan Pollock can be predicted automatically by taking the image of the fish and using it in one of the correlations developed in this study. The removal of the fins or the fins and the tail did not increase the prediction accuracy of the method. Therefore, intact fish images should be used.     

In situ activity of chymotrypsin in sugar‐salted herring during cold storage

In situ activity of intestinal chymotrypsin in sugar‐salted whole herring during cold storage was evaluated by analysing changes in the low‐molecular‐weight nitrogen fraction when a specific inhibitor was added. Addition of chymostatin (0.01 mM ) to sugar‐salted herring gave 100% inhibition of chymotrypsin activity compared to sugar‐salted herring without chymostatin. Inhibition of chymotrypsin did not affect the profile of low‐molecular‐weight peptides analysed by capillary zone electrophoresis, but resulted in 22% lower content after 26 weeks of storage. Inhibition of chymotrypsin resulted in 17% lower content of free amino acids. The relative amount of single free amino acids was unaffected by the presence of inhibitor, except for valine, leucine, isoleucine, methionine, phenylalanine and tyrosine, which were formed in lower relative amount. Removal of intestines (gutted herring) and thereby all intestinal proteolytic activity did not change the free amino acid profile, except for the concentration of histidine which decreased more when intestines were present. © 2000 Society of Chemical Industry     

Predicting intramuscular fat content and marbling score of pork along the longissimus muscle based on the last rib

The possibility of using IMF content/MS at the last rib to predict the corresponding fat attribute of loin chops at the 2nd last to the 7th last thoracic ribs and the mean of IMF content/MS of the whole loin joint was investigated. Linear and quadratic regression, cross‐validation analysis and analysis of variance were employed for the study. For either IMF or MS, linear and quadratic regression models had similar performance. The best mean correlation coefficients of calibration (R_c) of 0.91 and cross‐validation (R_v) of 0.88 were obtained for linear regression model of IMF content of the whole loin, while R_c of 0.90 and R_v of 0.88 were obtained for quadratic regression model of MS of the whole loin. The results indicated that IMF content and MS at the last rib had great potential for the prediction of IMF content and MS at the other ribs.     

Protein content determination in Brassica oleracea species using FT‐NIR technique and PLS regression

Six fresh and one frozen vegetable cultivar groups possessing remarkably different morphology from the same Brassica oleracea species, including broccoli, Brussels sprouts, curly cabbage, white cabbage, red cabbage, cauliflower and white kohlrabi, were chosen to set up a Fourier transform near‐infrared spectroscopy (FT‐NIR)‐based method for the quantification of protein content. Sample preparation was based on lyophilisation and homogenisation. Calibration was set up with the help of the Kjeldahl method for the quantification of protein content in the range of 12.9–32.5 m/m%. Calibration model was developed using the spectral regions 1136–1334 and 1639–1836 nm, with partial least squares regression. This model was checked by cross‐validation. The performance of the final FT‐NIR estimation model was evaluated by root mean square of cross‐validation, root‐mean‐square error of estimation and the determination coefficient (R²). The final estimation function for the protein determination was characterised with the predictive error of 0.76 m/m% and R² value of 98.81.     

Influence of salt concentration on properties of marinated meat from fresh and frozen herring (Clupea harengus L.)

Salt concentrations of brine above 10% are still commonly used in fish marinating process. The study has showed an increasing salt concentrations affect to adversely all the properties of marinated herring meat. Increase in salt concentration from 5% to 15% resulted in significantly (P < 0.05) decrease content of water, non‐protein nitrogen and products of protein hydrolysis. Consequently, weight yield of marinated fresh herring decreased from 83% to 74%. The most palatable marinades contained 2–3% NaCl in meat only. The higher salt concentrations applied significantly (P < 0.05) worsened the taste, texture and colour of marinated herring. Texture profile analyses (TPA), free hydroxyproline content and colour analyses (L*a*b) confirmed the negative influence of high salt concentration on the marinades quality. Frozen and thawed herring tissue showed greater sensitivity to salt. Namely, marinades from frozen fish contained by 0.3 more salt (P < 0.05), and the weight yield was by 2.3–10.3 percent point lower than from the fresh herring. Meat from frozen herring when matured has significantly (P < 0.05) lower sensory value, lower content of protein hydrolysis products, free hydroxyproline and higher parameter b* value of than the fresh one.     

基于IGS-SVM模型的牛肉生理成熟度预测方法

生理成熟度是判定牛肉质量等级的重要指标,本实验建立一种通过改进的网格搜索（improved grid search,IGS）算法优化支持向量机（support vector machine,SVM）参数的模型,以实现牛肉的生理成熟度的预测。收集18、36、54、72 月龄的牛肉样本各25 个,共计100 个。利用机器视觉,采集样本的显微图像,经过图像处理后,提取不同生理成熟度牛肉的肌纤维特征参数,用统计学方法分析牛肉生理成熟度和肌纤维特征参数之间的相关性,并以肌纤维特征参数作为输入,利用76 个训练集样本,建立牛肉生理成熟度的SVM预测模型。为优化所建立的SVM模型,提出一种IGS算法,用其对SVM模型的约束参数C以及核函数参数g进行优化,结合留一交叉验证法得到最优的（C,g）参数组合,并将最佳参数代入分类器,得到优化的牛肉生理成熟度预测模型。用24 个测试集的独立样本检测模型的适用性并估测性能,结果表明：利用该模型对牛肉生理成熟度预测的准确率可达到91.67%;与传统网格搜索算法相比,IGS算法使得模型在训练时间上缩短了1 755.41 s。这表明所建立的模型具有较好的预测效果,也说明根据牛肉肌纤维的特征参数结合机器视觉及图像分析技术,对牛肉生理成熟度进行自动判定的方法是可行的。     

Comparison of Minolta colorimeter and machine vision system in measuring colour of irradiated Atlantic salmon

BACKGROUND: Minolta and machine vision are two different instrumental techniques used for measuring the colour of muscle food products. Between these two techniques, machine vision has many advantages, such as its ability to determine L*, a*, b* values for each pixel of a sample's image and to analyse the entire surface of a food regardless of surface uniformity and colour variation. The objective of this study was to measure the colour of irradiated Atlantic salmon fillets using a hand‐held Minolta colorimeter and a machine vision system and to compare their performance. RESULTS: The L*, a*, b* values of Atlantic salmon fillets subjected to different electron beam doses (0, 1, 1.5, 2 and 3 kGy) were measured using a Minolta CR‐200 Chroma Meter and a machine vision system. For both Minolta and machine vision the L* value increased and the a* and b* values decreased with increasing irradiation dose. However, the machine vision system showed significantly higher readings for L*, a*, b* values than the Minolta colorimeter. Because of this difference, colours that were actually measured by the two instruments were illustrated for visual comparison. Minolta readings resulted in a purplish colour based on average L*, a*, b* values, while machine vision readings resulted in an orange colour, which was expected for Atlantic salmon fillets. CONCLUSION: The Minolta colorimeter and the machine vision system were very close in reading the standard red plate with known L*, a*, b* values. Hence some caution is recommended in reporting colour values measured by Minolta, even when the ‘reference’ tiles are measured correctly. The reason for this discrepancy in colour readings for salmon is not known and needs further investigation. Copyright © 2009 Society of Chemical Industry     

Frozen herring as raw material for spice‐salting

One batch of herring (Clupea harengus) was spice‐salted fresh and as thawed after 32 days of frozen storage at −24 °C. After salting, samples of both groups were sent to participating laboratories in Iceland, Denmark, Norway, Germany and England. The herring was kept at 5 ± 1 °C and sampled three times during a 26 week storage period. The development of taste and texture characteristics (determined by sensory evaluation and instrumental texture measurements), formation of low‐molecular‐weight nitrogen compounds and changes in proteolytic activity were followed in both groups. The sensory evaluation results showed that thawed salted herring ripened in a similar manner to herring salted fresh, but at a faster rate. Instrumental texture analysis showed a faster rate of tenderising in thawed salted herring. Proteolytic activity measured as general activity and with specific synthetic substrates was higher in the thawed salted herring. The formation of trichloroacetic acid‐soluble nitrogen and free amino acids was faster in the previously frozen herring. The results show an accelerated but similar rate of ripening in thawed spice‐salted herring in comparison with fresh salted herring. © 2000 Society of Chemical Industry     

Development of a green two‐dimensional HPLC‐DAD/ESI‐MS method for the determination of anthocyanins from Prunus cerasifera var. atropurpurea leaf and improvement of their stability in energy drinks

This study aimed to develop a green two‐dimensional HPLC‐DAD/ESI‐MS method for analysing anthocyanins from Prunus cerasifera var. atropurpurea leaf and improve their stability in energy drinks by the addition of phenolic acids. Ethanol and tartaric acid solutions were used as mobile phases for one‐dimensional HPLC‐DAD for quantitative analysis of anthocyanins, and the primary anthocyanins were identified as cyanidin‐3‐O‐galactoside, cyanidin‐3‐O‐glucoside and cyanidin‐3‐O‐rutinoside using two‐dimensional HPLC‐MS. Method validation showed that the developed method was accurate, stable and reliable for the analysis of P. cerasifera anthocyanins. The effects of gallic, ferulic and caffeic acid on the stability of cyanidin‐3‐O‐galactoside, cyanidin‐3‐O‐glucoside, cyanidin‐3‐O‐rutinoside and total anthocyanins from P. cerasifera leaf in energy drinks were evaluated, and the degradation of P. cerasifera anthocyanins ideally followed a first‐order model (R² > 0.98). Gallic acid showed stronger protective effects on P. cerasifera anthocyanins in energy drinks, and adding/increasing ferulic and caffeic acids accelerated the degradation reactions.     

Using Image Analysis to Predict the Weight of Alaskan Salmon of Different Species

ABSTRACT: After harvesting, salmon is sorted by species, size, and quality. This is generally manually done by operators. Automation would bring repeatability, objectivity, and record-keeping capabilities to these tasks. Machine vision (MV) and image analysis have been used in sorting many agricultural products. Four salmon species were tested: pink (Oncorhynchus gorbuscha), red (Oncorhynchus nerka), silver (Oncorhynchus kisutch), and chum (Oncorhynchus keta). A total of 60 whole fish from each species were first weighed, then placed in a light box to take their picture. Weight compared with view area as well as length and width correlations were developed. In addition the effect of “hump” development (see text) of pink salmon on this correlation was investigated. It was possible to predict the weight of a salmon by view area, regardless of species, and regardless of the development of a hump for pinks. Within pink salmon there was a small but insignificant difference between predictive equations for the weight of “regular” fish and “humpy” fish. Machine vision can accurately predict the weight of whole salmon for sorting.     

Determination of fat in live farmed Atlantic salmon using non‐invasive NIR techniques

Non‐destructive near‐infrared (NIR) measurements were performed on 100 live, anaesthetised farmed Atlantic salmon, whole weight 1–11 kg, using two different NIR instruments: a grating monochromator instrument equipped with a fibre optic interactance probe, and a diode array instrument measuring diffuse reflectance in a non‐contact mode. Crude fat content was determined using partial least squares (PLS) regression. Full cross‐validation was used to evaluate the performance of the calibration models, expressed as the root mean square error of prediction (RMSEP). For the fibre optic instrument the wavelength range from 800 to 1098 nm resulted in a correlation coefficient of 0.90 and an RMSEP equal to 14 g kg^?1 fat. The diode array instrument using the wavelength range from 900 to 1700 nm gave results of the same accuracy. The measurement times were 21 and 3 s respectively. It is concluded that either instrument could be used to determine the crude fat content in live Atlantic salmon, with good accuracy. © 2003 Society of Chemical Industry     

SEGMENTATION OF BEEF JOINT IMAGES USING HISTOGRAM THRESHOLDING

Four histogram‐based thresholding methods, i.e., one‐dimensional (1‐D) histogram variance, 1‐D histogram entropy, two‐dimensional (2‐D) histogram variance and 2‐D histogram entropy, were proposed to segment the images of beef joints (raw, cooked and cooled) automatically from the background. The 2‐D histogram‐based methods incorporate a fast algorithm to reduce the calculation time, thus increasing the speed greatly. All the four methods were applied to 15 beef joint images captured from a video camera, and the methods including pixel classification, object overlap and object contrast for the evaluation of segmentation results were then employed to compare the performances or the abilities of the four different segmenting methods. Results indicate that the 2‐D histogram variance thresholding method can accomplish the segmentation task with the most satisfactory performance.     

梳棉机的发展现状及展望

探讨梳棉机发展现状及发展方向。对国内外新型梳棉机的生条定量、出条速度、锡林直径、幅宽、刺辊、棉网清洁器、盖板根数等进行了总结分析,指出由于锡林速度和机幅进一步加宽及弹性回转盖板系统等固有的限制,当今回转盖板梳棉机进一步增加产量将会十分困难。认为:开发高效能金属针布和重新设计回转盖板系统是梳棉机下一步提高产量的主攻方向;具有高效棉网清洁器和高效能固定盖板针布系统的全固定盖板梳棉机是梳棉机进一步提高产量的另一途径。     

Texture Quality Analysis of Rainbow Trout Using Hyperspectral Imaging Method

In this study a hyperspectral imaging system (short wave infrared range from 1000 to 2500 nm) was used to model fish texture by experimental compression test. Partial least square–discriminate analysis modeling technique was used for classifying the samples by linking the hyperspectral information and their measured texture. The R² of cross validation and prediction were 0.97 and 0.96, respectively. The root mean squared errors for cross validation and prediction were 0.07 and 0.09, respectively. Sensitivity and specificity for both class I and II were 1.00. Results indicated that hyperspectral imaging in short wave infrared range has ability to detect texture stiffness of rainbow trouts which is affected by freshness.     

A Deep Convolutional Neural Network Architecture for Boosting Image Discrimination Accuracy of Rice Species

In this paper, we proposed a machine vision system based on deep convolutional neural network (DCNN) architecture for improving the accuracy of classifying three distinct groups of rice kernel images compared with the traditional approaches. The main advantage of the presented method was able to avoid many heuristics as well as manual labor to tune complex parameters according to the domain to reach a modest level of accuracy in the classical feature extraction algorithms. We trained our models using stochastic gradient descent with momentum of 0.9 and weight decay of 0.0005 to optimize the network parameters and minimize the back-propagation error on the training dataset. We used a batch size between 15 and 150 and epochs time configured between 10 and 25. The experiment results showed that the highest accuracy of 99.4% obtained in the training process with batch size of 15 and epoch time of 20. We also compared the DCNN method with the traditional hand-engineered approaches of PHOG-KNN, PHOG-SVM, GIST-KNN, and GIST-SVM for rice kernel classification. The results showed that DCNN routinely outperforms other methods in similar machine vision tasks. The prediction accuracy results for test datasets by PHOG-KNN, PHOG-SVM, GIST-KNN, and GIST-SVM models were 89.1, 76.9, 90.6, and 92.1%, respectively. The highest prediction accuracy of DCNN is 95.5%, which showed the effectiveness of our proposed method for rice kernel classification. The aim of this study is to set up an automatic and accurate intelligent detection system and offering much value to current rice processing industry. With the comparably high classification accuracy, developed neural network could be used as a tool to achieve better and more objective rice quality evaluation at trading points within the rice marketing system.