Advanced deep learning approach for enhancing crop disease detection in agriculture using hyperspectral imaging

Article type: Research Article

Authors: Zekrifa, Djabeur Mohamed Seifeddine^{a; *} | Lamani, Dharmanna^b | Chaitanya, Gogineni Krishna^c | Kanimozhi, K.V.^d | Saraswat, Akash^e | Sugumar, D.^f | Vetrithangam, D.^g | Koshariya, Ashok Kumar^h | Manjunath, Manthur Sreeramulu^h | Rajaram, A.ⁱ

Affiliations: [a] Department of Bioinformatics, Higher School of Food Science and Agri-Food Industry, Ahmed Hamidouche Av, Oued Smar, Algiers, Algeria | [b] Department of Computer Science & Information Technology, School of Computing and Information Technology, REVA University, Rukmini Knowledge Park, Kattigenahalli, Bengaluru, Karnataka | [c] Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Andhra Pradesh, India | [d] Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Saveetha University, Chennai, Tamil Nadu, India | [e] Department of Computer Science Engineering, Dr. K. N. Modi University, Newai, Jaipur, Rajasthan | [f] Department of Electrical and Electronics Engineering, Karunya Institute of Technology and Sciences (Deemed to be University), Coimbatore, India | [g] Department of Computer Science and Engineering, University Institute of Engineering, Chandigarh University, Mohali, Punjab, India | [h] Department of Plant Pathology, School of Agriculture, Lovely Professional University, Phagwara, Punjab, India | [i] Department of Electronics and Communication Engineering, E.G.S. Pillay Engineering College, Nagapattinam, Tamilnadu, India

Correspondence: [*] Corresponding author. Djabeur Mohamed Seifeddine Zekrifa, Department of Bioinformatics, Higher School of Food Science and Agri-Food Industry, Ahmed Hamidouche Av, Oued Smar, Algiers, Algeria. E-mail: [email protected].

Abstract: Crop diseases pose significant challenges to global food security and agricultural sustainability. Timely and accurate disease detection is crucial for effective disease management and minimizing crop losses. In recent years, hyperspectral imaging has emerged as a promising technology for non-destructive and early disease detection in crops. This research paper presents an advanced deep learning approach for enhancing crop disease detection using hyperspectral imaging. The primary objective is to propose a hybrid Autoencoder-Generative Adversarial Network (AE-GAN) model that effectively extracts meaningful features from hyperspectral images and addresses the limitations of existing techniques. The hybrid AE-GAN model combines the strengths of the Autoencoder for feature extraction and the Generative Adversarial Network for synthetic sample generation. Through extensive evaluation, the proposed model outperforms existing techniques, achieving exceptional accuracy in crop disease detection. The results demonstrate the superiority of the hybrid AE-GAN model, offering substantial advantages in terms of feature extraction, synthetic sample generation, and utilization of spatial and spectral information. The proposed model’s contributions to sustainable agriculture and global food security make it a valuable tool for advancing agricultural practices and enhancing crop health monitoring. With its promising implications, the hybrid AE-GAN model represents a significant advancement in crop disease detection, paving the way for a more resilient and food-secure future.

Keywords: Autoencoder-generative adversarial network (AE-GAN)

DOI: 10.3233/JIFS-235582

Journal: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 2, pp. 3281-3294, 2024