Vegetable and Fruit Image Classification with SqueezeNet based Deep Feature Generator

Automatic classification of food products according to their types is one of the most common problems in computer vision. In this paper, 15 different types of vegetables were automatically classified through transfer learning in deep learning. The dataset used in the study is large and consists of 21,000 vegetable images. These images in the dataset are divided into 3 groups as training, testing and validation. Within the scope of the study, all of these groups were combined and a large dataset was obtained. SqueezeNet architecture is used for feature extraction in the developed deep learning-based machine learning model. In addition, the ReliefF method was used for feature selection and the most significant features were determined by eliminating negative features. In the classification phase of the developed application, Linear Discriminant Analysis (LDA) method was preferred. In this study, Hold-Out and 10-fold cross-validation techniques were used. Approximately 99% accuracy value was obtained in both validation techniques. The obtained results of the study show that the proposed method can be used successfully in automatic vegetable classification.

Keywords:

Vegetable classification, SqueezeNet, ReliefF Linear Discriminant Analysis,

PDF

___

G.J.H. Grubben, Vegetables, Prota, 2004.
J.L. Slavin, B. Lloyd, Health benefits of fruits and vegetables, Adv. Nutr. 3 (2012) 506–516.
J.C. Rickman, C.M. Bruhn, D.M. Barrett, Nutritional comparison of fresh, frozen, and canned fruits and vegetables II. Vitamin A and carotenoids, vitamin E, minerals and fiber, J. Sci. Food Agric. 87 (2007) 1185–1196.
K. Ikeuchi, Computer vision: A reference guide, Springer, 2021.
A.I. Khan, S. Al-Habsi, Machine learning in computer vision, Procedia Comput. Sci. 167 (2020) 1444–1451.
M. Hassaballah, K.M. Hosny, Recent advances in computer vision, Stud. Comput. Intell. 804 (2019).
J.F.S. Gomes, F.R. Leta, Applications of computer vision techniques in the agriculture and food industry: a review, Eur. Food Res. Technol. 235 (2012) 989–1000.
D. Wu, D.-W. Sun, Colour measurements by computer vision for food quality control–A review, Trends Food Sci. Technol. 29 (2013) 5–20.
S. Li, Y. Tian, P. Jiang, Y. Lin, X. Liu, H. Yang, Recent advances in the application of metabolomics for food safety control and food quality analyses, Crit. Rev. Food Sci. Nutr. 61 (2021) 1448–1469.
S. Gaikwad, Literature Review on Multi-Spectral Imaging for Fruits and Vegetable, Available SSRN 3905180. (2021).
F. Yuesheng, S. Jian, X. Fuxiang, B. Yang, Z. Xiang, G. Peng, W. Zhengtao, X. Shengqiao, Circular fruit and vegetable classification based on optimized GoogLeNet, IEEE Access. 9 (2021) 113599–113611.
J.K. Bhavya, B.S. AC, K. Gayithri, B.L. Keerthi, M.Y. MG, THE LITERATURE SURVEY ON INTRA CLASS FRUITS AND VEGETABLE RECOGNITION SYSTEM USING DEEP LEARNING, (n.d.).
R.S. Latha, G.R. Sreekanth, R.C. Suganthe, M. Geetha, N. Swathi, S. Vaishnavi, P. Sonasri, Automatic Fruit Detection System using Multilayer Deep Convolution Neural Network, in: 2021 Int. Conf. Comput. Commun. Informatics, IEEE, 2021: pp. 1–5.
M.I. Ahmed, S.M. Mamun, A.U.Z. Asif, DCNN-Based Vegetable Image Classification Using Transfer Learning: A Comparative Study, in: 2021 5th Int. Conf. Comput. Commun. Signal Process., IEEE, 2021: pp. 235–243.
O. Patil, Classification of Vegetables using TensorFlow, Int. J. Res. Appl. Sci. Eng. Technol. 6 (2018) 2926–2934. https://doi.org/10.22214/ijraset.2018.4488.
H. Kuang, C. Liu, L.L.H. Chan, H. Yan, Multi-class fruit detection based on image region selection and improved object proposals, Neurocomputing. 283 (2018) 241–255.
Z. Yuhui, C. Mengyao, C. Yuefen, L. Zhaoqian, L. Yao, L. Kedi, An Automatic Recognition Method of Fruits and Vegetables Based on Depthwise Separable Convolution Neural Network, J. Phys. Conf. Ser. 1871 (2021). https://doi.org/10.1088/1742-6596/1871/1/012075.
J.L. Joseph, V.A. Kumar, S.P. Mathew, Fruit Classification Using Deep Learning, in: Innov. Electr. Electron. Eng., Springer, 2021: pp. 807–817.
F.N. Iandola, S. Han, M.W. Moskewicz, K. Ashraf, W.J. Dally, K. Keutzer, SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size, ArXiv Prepr. ArXiv1602.07360. (2016).
B. Koonce, SqueezeNet, in: Convolutional Neural Networks with Swift Tensorflow, Springer, 2021: pp. 73–85.
T. Tuncer, E. Akbal, S. Dogan, Multileveled ternary pattern and iterative ReliefF based bird sound classification, Appl. Acoust. 176 (2021) 107866. https://doi.org/10.1016/j.apacoust.2020.107866.
P. Xanthopoulos, P.M. Pardalos, T.B. Trafalis, Linear discriminant analysis, in: Robust Data Min., Springer, 2013: pp. 27–33.
M.I. Ahmed, Vegetable Image Dataset, (2022). https://www.kaggle.com/misrakahmed/vegetable-image-dataset.
D.M.W. Powers, Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation, (2020). https://arxiv.org/abs/2010.16061 (accessed November 15, 2021).