Performance Analysis of LVQ and Euclidean Distance for Chili Type Detection: The Impact of Training Iterations on System Generalization
Main Article Content
Abstract
This research is driven by the challenges faced by farmers and consumers in Kerik Village, Magetan Regency, in accurately differentiating chili varieties, which impacts service quality and increases the risk of transactional errors. The study aims to implement and evaluate the effectiveness of an automated chili detection system by integrating the Learning Vector Quantization (LVQ) algorithm with the Euclidean Distance metric. The research methodology follows a structured flow, including the collection of primary digital images from local agricultural sites, grayscale pre-processing, and performance evaluation based on varying training iterations. The experimental results indicate that training intensity significantly influences the model's reliability. The system achieved its peak performance at 1,000 iterations, yielding a training accuracy of 71.67% and a superior testing accuracy of 85.00%. These findings demonstrate that the model possesses strong generalization capabilities without experiencing overfitting, outperforming several baseline agricultural classification studies. Practically, this automated tool provides a strategic solution for the farming community to mitigate misidentification risks, thereby enhancing product consistency and supporting localized smart farming initiatives.
Downloads
Article Details
Section
How to Cite
References
[1] A. Wulansari, “Metode pengujian akurasi pada klasifikasi citra digital,” Jurnal Sains dan Edukasi, vol. 5, no. 3, pp. 201–209, 2017.
[2] P. Utomo, “Klasifikasi jenis beras menggunakan jaringan saraf tiruan Learning Vector Quantization,” Jurnal Teknologi Informasi, vol. 11, no. 2, pp. 45–51, 2016.
[3] S. Soepomo, Budidaya Tanaman Hortikultura: Cabai dan Varietasnya. Penerbit Andi, 2013.
[4] M. S. Azhari and E. Supratman, “Klasifikasi jenis buah nanas menggunakan algoritma Learning Vector Quantization (LVQ),” Jurnal Informatika dan Komputer, vol. 12, no. 2, pp. 115–122, 2021.
[5] D. Gustina, M. F. Adiwisastra, and V. Lusiana, “Deteksi jenis bunga kamboja Bali berdasarkan pola daun menggunakan jaringan saraf tiruan Learning Vector Quantization,” J. Teknol., vol. 8, no. 1, pp. 22–29, 2016.
[6] S. R. Andani and A. S. Nugraha, “Mango variety classification using Learning Vector Quantization based on Canny edge detection and GLCM features,” Journal of Applied Informatics and Computing (JAIC), vol. 4, no. 1, pp. 45–52, 2020.
[7] T. Rizaldi, L. Hakim, and A. Wijaya, “Virtual fitting room implementation using Euclidean Distance for body measurement accuracy,” Journal of Computer Science and Engineering, vol. 5, no. 2, pp. 142–149, 2018.
[8] M. Jannah and N. H. Humaira, “Human gait recognition using Euclidean Distance and skeleton-based features,” International Journal of Artificial Intelligence & Robotics (IJAIR), vol. 1, no. 2, pp. 87–94, 2019.
[9] L. Munantri, M. Sobri, and S. Nurhayati, “Pengolahan citra digital untuk identifikasi ciri morfologi tanaman,” Jurnal Sistem Informasi dan Teknologi, vol. 2, no. 3, pp. 78–85, 2020.
[10] P. Utomo and B. D. Cahyono, “Penentuan masa panen padi berdasarkan warna daun menggunakan metode LVQ,” Jurnal Ilmiah Teknologi dan Informasijaya, vol. 3, no. 1, pp. 10–18, 2014.
[11] M. Sukri, R. Rahim, and K. Khairul, “Implementasi metode Euclidean Distance dalam sistem pengenalan objek,” Jurnal Riset Komputer (JURIKOM), vol. 3, no. 6, pp. 112–117, 2016.
[12] A. Wulansari, “Metode pengujian akurasi pada klasifikasi citra digital,” Jurnal Sains dan Edukasi, vol. 5, no. 3, pp. 201–209, 2017.
[13] M. S. Azhari and E. Supratman, “Klasifikasi jenis buah nanas menggunakan algoritma Learning Vector Quantization (LVQ),” Jurnal Informatika dan Komputer, vol. 12, no. 2, pp. 115–122, 2021.
[14] P. Utomo and D. N. Amadi, “Group Decision Making Using Mean SAW Borda and Decision Maker-Based Criteria Weighting,” Journal of Information Systems and Informatics, vol. 7, no. 2, Jun. 2025.