Table 4.
Applications of MVS with traditional ML in food processing.
| Products | Species | Application | Classification Methods | Evaluation | Reference |
|---|---|---|---|---|---|
| Animal | Beef | Predicting | a regression model |
R2 = 98.2,P< 0.05 adjusted |
Amani et al. (2015) |
| Clam | Detecting | Binary DT | accuracy = 98% | Coelho et al. (2016) | |
| Grading | SPA–PLSR | RP = 0.801, RMSEP = 0.157 | Xiong et al. (2015) | ||
| Chicken | Grading | PLSR | RMSEp, multiple results | Yang et al. (2018) | |
| Egg | Grading | SPA-SVR, SVC | 96.3% for scattered yolk | Zhang et al. (2015) | |
| Salmon | Grading | PLSR |
rcv = 0.834(driploss) rcv = 0.877(PH) |
He et al. (2014) | |
| Grading | TreeBagger | accuracy = 97.8% | Xu et al. (2016) | ||
| Fruit | Apple | Grading | RVM | accuracy = 95.63% | Zhang et al. (2014) |
| Grading | PLS, CARS | rp = 0.977,0.977,0.955 (three positions) | Fan et al. (2016a) | ||
| Grading | MLR | R = 0.90, RMSECV = 6.99N | Sun et al. (2016) | ||
| Grading | CPLS | r = 0.9327 | Fan et al. (2016b) | ||
| Grading | a bi-layer model | r = 0.9560 | Tian et al. (2017) | ||
| Grading | PLS | R2p = 0.83 | Khatiwada et al. (2016) | ||
| Grading | PLS-DA, PBR | accuracy = 98% | Keresztes et al. (2016) | ||
| Apricot | Grading | PLS | Büyükcan et al. (2016) | ||
| Blueberry | Grading | CARS-LS-SVM | accuracy = 93.3% (for healthy), accuracy = 98.0% (for bruised) | Fan et al. (2017) | |
| Grading | logistic function tree | accuracy = 95.2% | Hu et al. (2016) | ||
| Grading | SVM | accuracy = 97% | Leiva-Valenzuela et al. (2013) | ||
| Cherry | harvesting | Bayesian | accuracy = 89.6% | Amatya et al. (2015) | |
| Citrus | Detecting | Gaussian–Lorentzian | accuracy = 93.4% | Lorente et al. (2015) | |
| Mango | Grading | SVR, MADM | accuracy = 87%. | Nandi et al. (2016) | |
| Grading | Fuzzy classifier | accuracy = 89% | Naik et al. (2017) | ||
| Peach | Grading | SPA | accuracy = 100% | Sun et al. (2017) | |
| Grading | An improved watershed segmentation algorithm | accuracy = 96.5% (for bruised), accuracy = 97.5% (for sound) | Li et al. (2018) | ||
| Pear | Grading | SPA-SVM | accuracy = 93.3%, 96.7% | Hu et al. (2017) | |
| Pomegranate | Grading | PLS | r = 0.97 | Khodabakhshian et al. (2016) | |
| Strawberry | Grading | SVM | accuracy = 100% | Liu et al. (2014) | |
| Vegetable | Tomato | Grading | DSSAEs | accuracy = 95.5% | Iraji (2018) |
| Onion | Grading | SVMs | accuracy = 88.9% | Wang et al. (2015) | |
| Potato | Grading | LDA-MD for color | above 90% for 5 potato cultivars (color) |
Noordam et al. (2000) | |
| Others | Beans | Classifying | K-means and KNN | accuracy = 99.88% | Araújo et al. (2015) |
| Cheese | Grading | PLSR | R2 = 0.8321 | Barreto et al. (2018) | |
| Coffee bean | Grading | linear estimation models | R2 = 0.93 | Ramos et al. (2017) | |
| Cookie, Potato Chips |
Monitoring | non-destructive computer vision-based image analysis | R2 = 0.895 | Ataç Mogol et al. (2014) | |
| Dried food | Monitoring | PCA, FCM | Aghbashlo et al. (2014) | ||
| General | Detecting | GMM | multiple results | Einarsdóttir et al. (2016) | |
| Grain | Monitoring | SMK–LSSVM | accuracy = 98.13% | Liu et al. (2016) | |
| Olive | Grading | PLSR, PCA, LDA | multiple results | Fernández-Espinosa (2015) | |
| Olive oil | Grading | ANN, SVM, BN | accuracy = 100% with BN | Sanaeifar et al. (2018) | |
| Potato Chips | Detecting | SVM | accuracy = 94% | Dutta et al. (2015) | |
| Rice | Monitoring | Fuzzy logic | accuracy = 89.2% | Zareiforoush et al. (2016) | |
| Sesame | Grading | CARS-LS-SVM, CARS-LDA |
accuracy = 100% | Xie et al. (2014) | |
| Soybean | Grading | PLSR | multiple results | Huang et al. (2014) | |
| Spring rolls, minced meat | Detection | SDA | 5% error with 10-fold cross-validation | Einarsson et al. (2017) | |
| Tomato Juice | Grading | PLSR | R2 = 0.75 | Deak et al. (2015) | |
| Walnut | Grading | SVM | Tran et al. (2017) |