A Comprehensive Method for Assessing Meat Freshness Using Fusing Electronic Nose, Computer Vision, and Artificial Tactile Technologies

The traditional methods cannot be used to meet the requirements of rapid and objective detection of meat freshness. Electronic nose (E-Nose), computer vision (CV), and artificial tactile (AT) sensory technologies can be used to mimic humans’ compressive sensory functions of smell, look, and touch when making judgement of meat quality (freshness). Though individual E-Nose, CV, and AT sensory technologies have been used to detect the meat freshness, the detection results vary and are not reliable. In this paper, a new method has been proposed through the integration of E-Nose, CV, and AT sensory technologies for capturing comprehensive meat freshness parameters and the data fusion method for analysing the complicated data with different dimensions and units of six odour parameters of E-Nose, 9 colour parameters of CV, and 4 rubbery parameters of AT for effective meat freshness detection. The pork and chicken meats have been selected for a validation test. The total volatile base nitrogen (TVB-N) assays are used to define meat freshness as the standard criteria for validating the effectiveness of the proposed method. The principal component analysis (PCA) and support vector machine (SVM) are used as unsupervised and supervised pattern recognition methods to analyse the source data and the fusion data of the three instruments, respectively. The experimental and data analysis results show that compared to a single technology, the fusion of E-Nose, CV, and AT technologies significantly improves the detection performance of various freshness meat products. In addition, partial least squares (PLS) is used to construct TVB-N value prediction models, in which the fusion data is input. The root mean square error predictions (RMSEP) for the sample pork and chicken meats are 1.21 and 0.98, respectively, in which the coefficient of determination (R2) is 0.91 and 0.94. This means that the proposed method can be used to effectively detect meat freshness and the storage time (days).

Download Full-text

ELECTRONIC NOSE FOR ANCHOVY FRESHNESS MONITORING BASED ON SENSOR ARRAY AND PATTERN RECOGNITION METHODS: PRINCIPAL COMPONENTS ANALYSIS, LINEAR DISCRIMINANT ANALYSIS AND SUPPORT VECTOR MACHINE

International Journal of Computing ◽

10.47839/ijc.6.3.452 ◽

2014 ◽

pp. 61-67

Author(s):

A. Amari ◽

N. El Bari ◽

B. Bouchikhi

Keyword(s):

Support Vector Machine ◽

Discriminant Analysis ◽

Linear Discriminant Analysis ◽

Electronic Nose ◽

Tin Dioxide ◽

Principal Component ◽

Support Vector ◽

Current Pattern ◽

Linear Discriminant ◽

Pattern Recognition Methods

An electronic nose based system, which employs an array of six inexpensive commercial gas sensors based on tin dioxide (Figaro Engineering Inc., Japan), has been used to analyse the freshness states of anchovies. Fresh anchovies were stored in a refrigerator at 4 ± 1°C over a period of 15 days. Electronic nose measurements need no sample preparation and the results indicated that the spoilage process of anchovies could be followed by using this technique. Conductance responses of volatile compounds produced during storage of anchovy were monitored and the result were analysed by multivariate analysis methods. In this paper principal component analysis (PCA) and linear discriminant analysis (LDA) were used to investigate whether the electronic nose was able to distinguishing among different freshness states (fresh, moderated and non-fresh samples). The loadings analysis was used to identify the sensors responsible for discrimination in the current pattern file. Therefore, the support vector machines (SVM) method was applied to the new subset, with only the selected sensors, to confirm that a subset of a few sensors can be chosen to explain all the variance. The results obtained prove that the electronic nose can discriminate successfully different freshness state using LDA analysis. Some sensors have the highest influence in the current pattern file for electronic nose. Support vector machine (SVM) model, applied to the new subset of sensors show the good performance.

Download Full-text

Comparative Analysis of Hybrid Models for Prediction of BP Reactivity to Crossed Legs

Journal of Healthcare Engineering ◽

10.1155/2017/2187904 ◽

2017 ◽

Vol 2017 ◽

pp. 1-13 ◽

Cited By ~ 1

Author(s):

Gurmanik Kaur ◽

Ajat Shatru Arora ◽

Vijender Kumar Jain

Keyword(s):

Prediction Models ◽

Fuzzy Inference ◽

Research Work ◽

Principal Component ◽

Coefficient Of Determination ◽

Percentage Error ◽

Support Vector ◽

Inference System ◽

Forward Stepwise ◽

Artificial Neural Network Ann

Crossing the legs at the knees, during BP measurement, is one of the several physiological stimuli that considerably influence the accuracy of BP measurements. Therefore, it is paramount to develop an appropriate prediction model for interpreting influence of crossed legs on BP. This research work described the use of principal component analysis- (PCA-) fused forward stepwise regression (FSWR), artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS), and least squares support vector machine (LS-SVM) models for prediction of BP reactivity to crossed legs among the normotensive and hypertensive participants. The evaluation of the performance of the proposed prediction models using appropriate statistical indices showed that the PCA-based LS-SVM (PCA-LS-SVM) model has the highest prediction accuracy with coefficient of determination (R2) = 93.16%, root mean square error (RMSE) = 0.27, and mean absolute percentage error (MAPE) = 5.71 for SBP prediction in normotensive subjects. Furthermore, R2 = 96.46%, RMSE = 0.19, and MAPE = 1.76 for SBP prediction and R2 = 95.44%, RMSE = 0.21, and MAPE = 2.78 for DBP prediction in hypertensive subjects using the PCA-LSSVM model. This assessment presents the importance and advantages posed by hybrid computing models for the prediction of variables in biomedical research studies.

Download Full-text

Discrimination between Alternative Herbal Medicines from Different Categories with the Electronic Nose

Sensors ◽

10.3390/s18092936 ◽

2018 ◽

Vol 18 (9) ◽

pp. 2936 ◽

Cited By ~ 6

Author(s):

Xianghao Zhan ◽

Xiaoqing Guan ◽

Rumeng Wu ◽

Zhan Wang ◽

You Wang ◽

...

Keyword(s):

Herbal Medicine ◽

Electronic Nose ◽

Nearest Neighbor ◽

Herbal Medicines ◽

Principal Component ◽

Support Vector ◽

Linear Discriminant ◽

Convenient Means ◽

Electronic Nose System ◽

Insight Into

As alternative herbal medicine gains soar in popularity around the world, it is necessary to apply a fast and convenient means for classifying and evaluating herbal medicines. In this work, an electronic nose system with seven classification algorithms is used to discriminate between 12 categories of herbal medicines. The results show that these herbal medicines can be successfully classified, with support vector machine (SVM) and linear discriminant analysis (LDA) outperforming other algorithms in terms of accuracy. When principal component analysis (PCA) is used to lower the number of dimensions, the time cost for classification can be reduced while the data is visualized. Afterwards, conformal predictions based on 1NN (1-Nearest Neighbor) and 3NN (3-Nearest Neighbor) (CP-1NN and CP-3NN) are introduced. CP-1NN and CP-3NN provide additional, yet significant and reliable, information by giving the confidence and credibility associated with each prediction without sacrificing of accuracy. This research provides insight into the construction of a herbal medicine flavor library and gives methods and reference for future works.

Download Full-text

Computer Vision Intelligent Approaches to Extract Human Pose and Its Activity from Image Sequences

Electronics ◽

10.3390/electronics9010159 ◽

2020 ◽

Vol 9 (1) ◽

pp. 159

Author(s):

Paulo J. S. Gonçalves ◽

Bernardo Lourenço ◽

Samuel Santos ◽

Rodolphe Barlogis ◽

Alexandre Misson

Keyword(s):

Computer Vision ◽

Human Body ◽

Prediction Models ◽

Image Sequences ◽

Video Frame ◽

Support Vector ◽

Training Phase ◽

Human Body Model ◽

Body Model ◽

Human Pose

The purpose of this work is to develop computational intelligence models based on neural networks (NN), fuzzy models (FM), support vector machines (SVM) and long short-term memory networks (LSTM) to predict human pose and activity from image sequences, based on computer vision approaches to gather the required features. To obtain the human pose semantics (output classes), based on a set of 3D points that describe the human body model (the input variables of the predictive model), prediction models were obtained from the acquired data, for example, video images. In the same way, to predict the semantics of the atomic activities that compose an activity, based again in the human body model extracted at each video frame, prediction models were learned using LSTM networks. In both cases the best learned models were implemented in an application to test the systems. The SVM model obtained 95.97% of correct classification of the six different human poses tackled in this work, during tests in different situations from the training phase. The implemented LSTM learned model achieved an overall accuracy of 88%, during tests in different situations from the training phase. These results demonstrate the validity of both approaches to predict human pose and activity from image sequences. Moreover, the system is capable of obtaining the atomic activities and quantifying the time interval in which each activity takes place.

Download Full-text

Discrimination of Volatiles of Shiitakes (Lentinula edodes) Produced during Drying Process by Electronic Nose

International Journal of Food Engineering ◽

10.1515/ijfe-2019-0233 ◽

2020 ◽

Vol 16 (1-2) ◽

Author(s):

Hui Zhang ◽

Jing Peng ◽

Yu-ren Zhang ◽

Qiang Liu ◽

Lei-qing Pan ◽

...

Keyword(s):

Electronic Nose ◽

Lentinula Edodes ◽

Solid Phase ◽

Principal Component ◽

Gas Chromatography Mass Spectrometry ◽

Support Vector ◽

Drying Process ◽

Drying Time ◽

Discriminant Models ◽

Accuracy Rates

AbstractThis study aimed to investigate the potential of electronic nose (E-nose) to differentiate volatiles of shiitakes produced at different drying stages. Shiitakes at different drying time slots were categorized into four groups (fresh, early, middle and late stage) by sensory evaluation. E-nose was used to analyze the volatiles and compared with headspace solid phase micro-extraction combined with gas chromatography-mass spectrometry (HS/GC-MS). The principal component analysis results showed that shiitakes at each stage could be successfully discriminated by E-nose and HS/GC-MS. The differences in volatile organic compounds produced at each stage were mainly caused by sulfurs and alcohols, leading to apparent changes of sensors sensitive to sulfurs, alcohols and aromatic compounds. The discriminant models were established by partial least squares discriminant analysis and support vector machine classification, with accuracy rates of 91.25 % and 95.83 %, respectively. The results demonstrated the potential use of E-nose in classifying and monitoring shiitakes during drying process.

Download Full-text

Support vector regression is an improvement for principle component analysis

Journal of University of Shanghai for Science and Technology ◽

10.51201/jusst/21/06462 ◽

2021 ◽

Vol 23 (06) ◽

pp. 1699-1715

Author(s):

Mohamed, A. M. ◽

◽

Abdel Latif, S. H ◽

Alwan, A. S. ◽

◽

...

Keyword(s):

Principal Component Analysis ◽

Support Vector Regression ◽

Principle Component Analysis ◽

Mean Squared Error ◽

Principal Component ◽

Component Analysis ◽

Kernel Functions ◽

Coefficient Of Determination ◽

Support Vector ◽

Principle Component

The principle component analysis is used more frequently as a variables reduction technique. And recently, an evolving group of studies makes use of machine learning regression algorithms to improve the estimation of empirical models. One of the most frequently used machines learning regression models is support vector regression with various kernel functions. However, an ensemble of support vector regression and principal component analysis is also possible. So, this paper aims to investigate the competence of support vector regression techniques after performing principal component analysis to explore the possibility of reducing data and having more accurate estimations. Some new proposals are introduced and the behavior of two different models 𝜀𝜀-SVR and 𝑣𝑣-SVR are compared through an extensive simulation study under four different kernel functions; linear, radial, polynomial, and sigmoid kernel functions, with different sample sizes, ranges from small, moderate to large. The models are compared with their counterparts in terms of coefficient of determination (𝑅𝑅2 ) and root mean squared error (RMSE). The comparative results show that applying SVR after PCA models improve the results in terms of SV numbers between 30% and 60% on average and it can be applied with real data. In addition, the linear kernel function gave the best values rather than other kernel functions and the sigmoid kernel gave the worst values. Under 𝜀𝜀-SVR the results improved which did not happen with 𝑣𝑣-SVR. It is also drawn that, RMSE values decreased with increasing sample size.

Download Full-text

Electronic Nose-Based Technique for Rapid Detection and Recognition of Moldy Apples

10.20944/preprints201903.0008.v1 ◽

2019 ◽

Author(s):

Wenshen Jia ◽

Gang Liang ◽

Hui Tian ◽

Jing Sun ◽

Cihui Wan

Keyword(s):

Neural Network ◽

Pattern Recognition ◽

Electronic Nose ◽

Gas Sensors ◽

Penicillium Expansum ◽

Support Vector ◽

Linear Discriminant ◽

Pattern Recognition Methods ◽

Group A ◽

Group B

In this paper, PEN3 electronic nose was used to detect and recognize fresh and moldy apples (inoculated with Penicillium expansum and Aspergillusniger) taken Golden Delicious apples as model subject. Firstly, the apples were divided into two groups: apples only inoculated with different molds (Group A) and mixed apples of inoculated apples with fresh apples (Group B). Then the characteristic gas sensors of the PEN3 electronic nose that were most closely correlated with the flavor information of the moldy apples were optimized and determined, which can simplify the analysis process and improve the accuracy of results. Four pattern recognition methods, including linear discriminant analysis (LDA), backpropagation neural network (BPNN), support vector machines (SVM) and radial basis function neural network (RBFNN), were then applied to analyze the data obtained from the characteristic sensors, respectively, aiming at establishing the prediction model of flavor information and fresh/moldy apples. The results showed that only the gas sensors of W1S, W2S, W5S, W1W and W2W in the PEN3 electronic nose exhibited strong signal response to the flavor information, indicating were most closely correlated with the characteristic flavor of apples and thus the data obtained from these characteristic sensors was used for modeling. The results of the four pattern recognition methods showed that BPNN presented the best prediction performance for the training and validation sets for both the Group A and Group B, with prediction accuracies of 96.29% and 90.00% (Group A), 77.70% and 72.00% (Group B), respectively. Therefore, it first demonstrated that PEN3 electronic nose can not only effectively detect and recognize the fresh and moldy apples, but also can distinguish apples inoculated with different molds.

Download Full-text

Retrieval and Mapping of Soil Organic Carbon Using Sentinel-2A Spectral Images from Bare Cropland in Autumn

Remote Sensing ◽

10.3390/rs13061072 ◽

2021 ◽

Vol 13 (6) ◽

pp. 1072

Author(s):

Ke Wang ◽

Yanbing Qi ◽

Wenjing Guo ◽

Jielin Zhang ◽

Qingrui Chang

Keyword(s):

Remote Sensing ◽

Organic Carbon ◽

Soil Organic Carbon ◽

Soil Properties ◽

Spectral Reflectance ◽

Prediction Models ◽

Coefficient Of Determination ◽

Support Vector ◽

Fertility Level ◽

Sentinel 2A

Soil is the largest carbon reservoir on the terrestrial surface. Soil organic carbon (SOC) not only regulates global climate change, but also indicates soil fertility level in croplands. SOC prediction based on remote sensing images has generated great interest in the research field of digital soil mapping. The short revisiting time and wide spectral bands available from Sentinel-2A (S2A) remote sensing data can provide a useful data resource for soil property prediction. However, dense soil surface coverage reduces the direct relationship between soil properties and S2A spectral reflectance such that it is difficult to achieve a successful SOC prediction model. Observations of bare cropland in autumn provide the possibility to establish accurate SOC retrieval models using the S2A super-spectral reflectance. Therefore, in this study, we collected 225 topsoil samples from bare cropland in autumn and measured the SOC content. We also obtained S2A spectral images of the western Guanzhong Plain, China. We established four SOC prediction models, including random forest (RF), support vector machine (SVM), partial least-squares regression (PLSR), and artificial neural network (ANN) based on 15 variables retrieved from the S2A images, and compared the prediction accuracy using RMSE (root mean square error), R2 (coefficient of determination), and RPD (ratio of performance to deviation). Based on the optimal model, the spatial distribution of SOC was mapped and analyzed. The results indicated that the inversion model with the RF algorithm achieved the highest accuracy, with an R2 of 0.8581, RPD of 2.1313, and RMSE of 1.07. The variables retrieved from the shortwave infrared (SWIR) bands (B11 and B12) usually had higher variable importance, except for the ANN model. SOC content mapped with the RF model gradually decreased with increasing distance from the Wei river, and values were higher in the west than in the east. These results matched the SOC distribution based on measurements at the sample sites. This research provides evidence that soil properties such as SOC can be retrieved and spatially mapped based on S2A images that are obtained from bare cropland in autumn.

Download Full-text

Evaluation of IoT-Enabled Monitoring and Electronic Nose Spoilage Detection for Salmon Freshness During Cold Storage

Foods ◽

10.3390/foods9111579 ◽

2020 ◽

Vol 9 (11) ◽

pp. 1579

Author(s):

Huanhuan Feng ◽

Mengjie Zhang ◽

Pengfei Liu ◽

Yiliu Liu ◽

Xiaoshuan Zhang

Keyword(s):

Cold Storage ◽

Electronic Nose ◽

Principal Component ◽

Storage Condition ◽

Quality Parameters ◽

Training Dataset ◽

Support Vector ◽

Accuracy Rate ◽

Specific Storage ◽

Test Dataset

Salmon is a highly perishable food due to temperature, pH, odor, and texture changes during cold storage. Intelligent monitoring and spoilage rapid detection are effective approaches to improve freshness. The aim of this work was an evaluation of IoT-enabled monitoring system (IoTMS) and electronic nose spoilage detection for quality parameters changes and freshness under cold storage conditions. The salmon samples were analyzed and divided into three groups in an incubator set at 0 °C, 4 °C, and 6 °C. The quality parameters, i.e., texture, color, sensory, and pH changes, were measured and evaluated at different temperatures after 0, 3, 6, 9, 12, and 14 days of cold storage. The principal component analysis (PCA) algorithm can be used to cluster electronic nose information. Furthermore, a Convolutional Neural Networks and Support Vector Machine (CNN-SVM) based algorithm is used to cluster the freshness level of salmon samples stored in a specific storage condition. In the tested samples, the results show that the training dataset of freshness is about 95.6%, and the accuracy rate of the test dataset is 93.8%. For the training dataset of corruption, the accuracy rate is about 91.4%, and the accuracy rate of the test dataset is 90.5%. The overall accuracy rate is more than 90%. This work could help to reduce quality loss during salmon cold storage.

Download Full-text

Development and application of a new low cost electronic nose for the ripeness monitoring of banana using computational techniques (PCA, LDA, SIMCA and SVM)

Czech Journal of Food Sciences ◽

10.17221/113/2014-cjfs ◽

2014 ◽

Vol 32 (No. 6) ◽

pp. 538-548 ◽

Cited By ~ 21

Author(s):

A. Sanaeifar ◽

S.S. Mohtasebi ◽

M. Ghasemi-Varnamkhasti ◽

H. Ahmadi ◽

J. Lozano

Keyword(s):

Electronic Nose ◽

Low Cost ◽

Principal Component ◽

Oxide Semiconductor ◽

Support Vector ◽

Computational Techniques ◽

Laboratory Equipment ◽

Linear Discriminant ◽

Ripening Stages ◽

Non Destructive

Potential application of a metal oxide semiconductor based electronic nose (e-nose) as a non-destructive instrument for monitoring the change in volatile production of banana during the ripening process was studied. The proposed e-nose does not need any advanced or expensive laboratory equipment and proved to be reliable in recording meaningful differences between ripening stages. Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Soft Independent Modelling of Class Analogy (SIMCA) and Support Vector Machines (SVM) techniques were used for this purpose. Results showed that the proposed e-nose can distinguish between different ripening stages. The e-nose was able to detect a clear difference in the aroma fingerprint of banana when using SVM analysis compared with PCA and LDA, SIMCA analysis. Using SVM analysis, it was possible to differentiate and to classify the different banana ripening stages, and this method was able to classify 98.66% of the total samples in each respective group. Sensor array capabilities in the classification of ripening stages using loading analysis and SVM and SIMCA were also investigated, which leads to develop the application of a specific e-nose system by applying the most effective sensors or ignoring the redundant sensors.  

Download Full-text