Evaluation of CNN based on Hyperparameters to Detect the Quality of Apples

Evaluation of CNN based on Hyperparameters to Detect the Quality of Apples

© 2022 by IJETT Journal
Volume-70 Issue-10
Year of Publication : 2022
Authors : Deepali M Bongulwar, Vijay Prakash Singh, S. N. Talbar
DOI : 10.14445/22315381/IJETT-V70I10P222

How to Cite?

Deepali M Bongulwar, Vijay Prakash Singh, S. N. Talbar, "Evaluation of CNN based on Hyperparameters to Detect the Quality of Apples," International Journal of Engineering Trends and Technology, vol. 70, no. 10, pp. 232-246, 2022. Crossref, https://doi.org/10.14445/22315381/IJETT-V70I10P222

It is essential for the food industry that fresh goods are automatically categorized. Various kinds of fruits are in the market, making it difficult to classify them according to quality. Manual sorting and evaluation of agricultural goods are possible, but it is not definitive, time-consuming, subjective, costly, and environmentally sensitive. Thus, fast, accurate, effective and automatic methods need to be introduced for inspecting the quality and grading of fruit. Classification of the quality of fruit and, as a result, gradation is critical in the industry for the development of good quality food products and the highest grade fruits that can be offered in the market. This research develops an automatic fruit grading system to grade apples based on their external qualities. The flaws on the fruit's peel have been used to determine if the apple is fresh/good or rotten/bad. It has been demonstrated that convolutional neural networks (CNNs) are efficient in several agricultural applications. Therefore, CNN architecture is utilized to build and train the classification model. This study's objective is to determine the effectiveness of the proposed CNN model considering four hyperparameters like an optimizer, learning rate, number of epochs and batchsize for determining the quality of apples. The two benchmark datasets, 'Fruits Fresh and Rotten' (Dataset1) and 'FruitsGB' (Fruits Good/Bad) (Dataset2), are employed to analyze the performance of the model. Dataset1 consists of 1,693 fresh and 2,342 rotten apples, and dataset2 consists of 1000 good and 1000 bad apple images. The accuracy and computation time are utilized for the evaluation of the classification performance of the proposed CNN model. Initially, the model's accuracy is improved by changing batchsize and keeping hyperparameters like epochs constant, and the best results for each optimizer and all learning rates are found. The batchsize that produced the best results is chosen, and the model is reassessed by adjusting the number of epochs, optimizer, and learning rates. Finally, the best outcomes are obtained. The presented model has achieved 100% accuracy on dataset2 with the optimizers SGDM and ADAM and 99.31% and 99.70% for dataset1 with SGDM and ADAM, respectively. The results reveal that the model's accuracy lowers with the increase in the learning rate, and adding more epochs does not improve the accuracy. The performance categorization of the model is assessed using additional metrics like Precision, Recall, F1 and F2 score, MCC, and AUC. Thus, a score of 100% is achieved for all these metrics.The optimizer SGDM gives more good results than ADAM and optimizes faster. The acquired experimental findings show that the accuracy of the proposed model relies not only on the hyperparameters but also on the dataset used and its size.

Convolutional Neural Network, Deep Learning, Fruit quality, Fruits Fresh and Rotten dataset, Hyperparameters.

[1] Bhargava, Anuja, and Atul Bansal, “Machine Learning-Based Quality Evaluation of Mono-Coloured Apples,” Multimedia Tools and Applications vol. 79, no. 31, pp. 22989-23006, 2020.
[2] Nikhitha, M., S. Roopa Sri, and B. Uma Maheswari, “Fruit Recognition and Grade of Disease Detection Using Inception V3 Model", 3rd International Conference on Electronics, Communication and Aerospace Technology (ICECA), IEEE, pp. 1040-1043, 2019.
[3] Al-Shawwa, Mohammed O, “Classification of Apple Fruits By Deep Learning", International Journal of Academic Engineering Research(IJAER), vol. 3, no. 12, 2020.
[4] Khan, Asif Iqbal, S. M. K. Quadri, and Saba Banday, “Deep Learning for Apple Diseases: Classification and Identification,” International Journal of Computational Intelligence Studies vol. 10, no. 1, pp. 1-12, 2021
[5] Magsi, Aurangzeb, J. Ahmed Mahar, and Shahid Hussain Danwar, “Date Fruit Recognition Using Feature Extraction Techniques and Deep Convolutional Neural Network,” Indian Journal of Science and Technology vol.12, no. 32, pp. 1-12, 2019.
[6] Da Costa, Arthur Z., Hugo EH Figueroa, and Juliana A. Fracarolli, “Computer Vision-Based Detection of External Defects on Tomatoes Using Deep Learning,” Biosystems Engineering vol. 190, pp. 131-144, 2020.
[7] Singh, Uday Pratap, Siddharth Singh Chouhan, Sukirty Jain, and Sanjeev Jain, “Multilayer Convolution Neural Network for the Classification of Mango Leaves Infected By Anthracnose Disease,” IEEE Access vol. 7, pp. 43721- 43729, 2019.
[8] Howlader, Md Rasel, Umme Habiba, Rahat Hossain Faisal, and Md Mostafijur Rahman, “Automatic Recognition of Guava Leaf Diseases Using Deep Convolution Neural Network", International Conference on Electrical, Computer and Communication Engineering (ECCE), pp. 1-5. IEEE, 2019.
[9] Militante, Sammy V., Bobby D. Gerardo, and Nanette V. Dionisio, “Plant Leaf Detection and Disease Recognition Using Deep Learning", IEEE Eurasia Conference on Iot, Communication and Engineering (ECICE), pp. 579-582. IEEE, 2019.
[10] Shruthi, U., V. Nagaveni, and B. K. Raghavendra, “A Review on Machine Learning Classification Techniques for Plant Disease Detection", International Conference on Advanced Computing & Communication Systems (ICACCS), pp. 281-284. IEEE, 2019.
[11] Tiwari, Divyansh, Mritunjay Ashish, Nitish Gangwar, Abhishek Sharma, Suhanshu Patel, and Suyash Bhardwaj, “Potato Leaf Diseases Detection Using Deep Learning", 4th International Conference on Intelligent Computing and Control Systems (ICICCS), IEEE, pp. 461-466, 2020.
[12] Turkoglu, Muammer, Davut Hanbay, and Abdulkadir Sengur, “Multi-Model LSTM-Based Convolutional Neural Networks for Detection of Apple Diseases and Pests,” Journal of Ambient Intelligence and Humanized Computing, pp.1-11, 2019.
[13] Roy, Kyamelia, Sheli Sinha Chaudhuri, and Sayan Pramanik, “Deep Learning Based Real-Time Industrial Framework for Rotten and Fresh Fruit Detection Using Semantic Segmentation,” Microsystem Technologies, vol.27, no. 9, pp. 3365-3375, 2021.
[14] Siddiqi, Raheel, “Automated Apple Defect Detection Using State-of-the-Art Object Detection Techniques,” SN Applied Sciences, vol. 1, no. 11, pp.1 -12, 2019.
[15] Bade Ashwini Vivekanand, M. Suresh Kumar, "Deep Learning Based Tomato PLDD,” International Journal of Engineering Trends and Technology, vol.70. no. 7 pp. 414- 421, 2022. Crossref, https://doi.org/10.14445/22315381/IJETT-V70I7P243.
[16] Palakodati, S.S.S., Chirra, V.R.R., Yakobu, D. and Bulla, S., “Fresh and Rotten Fruits Classification Using CNN and Transfer Learning”, Revue d'Intelligence Artificielle (RIA), vol.34, no. 5, pp. 617-622, 2020.
[17] Alsayed, Ashwaq, Amani Alsabei, and Muhammad Arif, "Classification of Apple Tree Leaves Diseases Using Deep Learning Methods,” International Journal of Computer Science & Network Security vol. 21, no. 7, pp. 324-330, 2021.
[18] Pérez-Pérez, Blanca Dalila, Juan Pablo Garcia Vazquez, and Ricardo Salomón-Torres, "Evaluation of Convolutional Neural Networks' Hyperparameters with Transfer Learning to Determine Sorting of Ripe Medjool Dates", Agriculture, vol. 11, no .2, pp. 115, 2021.
[19] Vijayakumar, T., and Mr R. Vinothkanna, “Mellowness Detection of Dragon Fruit Using Deep Learning Strategy,” Journal of Innovative Image Processing (JIIP) vol. 2, no. 01, pp. 35-43, 2020.
[20] Kaggle Dataset Link, [Online]. Available: https://www.kaggle.com/Datasets/Sriramr/Fruits-Fresh-and-Rotten-for-Classification
[21] IEEE Dataport Dataset Link, [Online]. Available: https://IEEE-Dataport.org/Open-Access/Fruitsgb-Top-Indian-Fruits-Quality
[22] Challa, Sravan Kumar, Akhilesh Kumar, and Vijay Bhaskar Semwal, “A Multi-Branch CNN-Bilstm Model for Human Activity Recognition Using Wearable Sensor Data", The Visual Computer, pp. 1-15, 2021.
[23] Wang, Hongjun, Et Al, “Research on Detection Technology of Various Fruit Disease Spots Based on Maskr-CNN", International Conference on Mechatronics and Automation (ICMA), IEEE, 2020.
[24] Karakaya, Diclehan, Oguzhan Ulucan, and Mehmet Turkan, “A Comparative Analysis on Fruit Freshness Classification,” Innovations in Intelligent Systems and Applications Conference (ASYU), IEEE, 2019.
[25] Alharbi, Asmaa Ghazi, and Muhammad Arif, “Detection and Classification of Apple Diseases Using Convolutional Neural Networks", 2nd International Conference on Computer and Information Sciences (ICCIS), IEEE, 2020.