Analisis Pengaruh Integrasi Fitur Visual dan Sensor Gas terhadap Klasifikasi Grade Mangga Arumanis
DOI:
https://doi.org/10.29407/54nvme92Abstract
Proses grading mangga secara manual masih berpotensi menghasilkan penilaian yang subjektif dan kurang konsisten karena bergantung pada pengalaman petugas. Penelitian ini mengusulkan sistem klasifikasi grade mangga Arumanis berbasis pendekatan multimodal yang mengintegrasikan fitur area, hue median, delta gas dari sensor MQ-135, serta fitur visual hasil ekstraksi MobileNetV2 yang direduksi menggunakan Principal Component Analysis (PCA). Seluruh fitur digunakan sebagai masukan model Bayesian Neural Network (BNN) untuk menghasilkan prediksi grade dan nilai uncertainty. Dataset terdiri atas 165 citra yang diperoleh dari 55 buah mangga Arumanis dengan tiga sudut pengambilan gambar. Hasil pengujian menunjukkan bahwa model usulan memperoleh akurasi 93,75%, lebih tinggi dibandingkan model MobileNetV2 sebesar 60,72% dan BNN berbasis fitur numerik sebesar 74,54%. Evaluasi menggunakan K-Fold Cross Validation menghasilkan rata-rata akurasi 93,75% dengan standar deviasi 7,81%. Hasil penelitian menunjukkan bahwa integrasi fitur visual dan sensor gas mampu meningkatkan performa klasifikasi grade mangga secara signifikan.
Keywords:
Bayesian Neural Network, grading mangga, MobileNetV2, multimodal, MQ-135##plugins.themes.default.displayStats.downloads##
References
[1] K. Fulgoni and V. L. Fulgoni, “Mango Consumption Was Associated with Higher Nutrient
Intake and Diet Quality in Women of Childbearing Age and Older Adults,” Nutrients, vol.
16, no. 2, Jan. 2024, doi: 10.3390/NU16020303.
[2] W. Nong et al., “Effects of Harvest Maturity on Microbial Community Composition,
Enzyme Activities, and Phenolic Acids in Upper Tobacco Leaves After Curing,” 2025.
[3] A. Martínez, A. Hernández, P. Arroyo, J. Lozano, M. de G. Córdoba, and A. Martín,
“Early Detection of Monilinia laxa in Yellow-Fleshed Peach Using a Non-Destructive ENose Approach,” Foods, vol. 14, no. 18, p. 3155, Sep. 2025, doi: 10.3390/foods14183155.
[4] Y. H. Roh et al., “Physiological Responses and Determination of Harvest Maturity in
‘Daehong’ Peach According to Days After Full Bloom,” Horticulturae, vol. 11, no. 9, p.
1013, Aug. 2025, doi: 10.3390/horticulturae11091013.
[5] R. Nithya, B. Santhi, R. Manikandan, M. Rahimi, and A. H. Gandomi, “Computer Vision
System for Mango Fruit Defect Detection Using Deep Convolutional Neural Network,”
Foods, vol. 11, no. 21, p. 3483, Nov. 2022, doi: 10.3390/foods11213483.
[6] W. Zaman, M. F. Siddique, S. U. Khan, and J.-M. Kim, “A New Dual-Input CNN for
Multimodal Fault Classification Using Acoustic Emission and Vibration Signals,” Eng.
Fail. Anal., vol. 179, p. 109787, Sep. 2025, doi: 10.1016/j.engfailanal.2025.109787.
[7] L. Colaco and P. Kamat, “Artificial Intelligence Advances for Cashew Fruit Maturity and
Quality Detection: A Systematic Review on Models, Sensors, and Farming Applications,”
J. Big Data, vol. 12, no. 1, p. 250, Nov. 2025, doi: 10.1186/s40537-025-01296-2.
[8] Y. M. Hirimutugoda, T. P. Silva, and N. M. Wagarachchi, “Handling the predictive
uncertainty of convolutional neural network in medical image analysis: a review,” J. Med.
Artif. Intell., vol. 6, no. 0, Oct. 2023, doi: 10.21037/JMAI-23-40/COIF.
[9] A. Luque, M. Mazzoleni, A. Carrasco, and A. Ferramosca, “Visualizing Classification
Results: Confusion Star and Confusion Gear,” IEEE Access, vol. 10, pp. 1659–1677, 2022,
doi: 10.1109/ACCESS.2021.3137630.
[10] L. Zhao and D. Um, “Precision Agriculture Based on Bayesian Neural Network”.
[11] F. J. Diaz Blasco et al., “Employment of MQ gas sensors for the classification of Cistus
ladanifer essential oils,” Microchemical Journal, vol. 206, p. 111585, Nov. 2024, doi:
10.1016/j.microc.2024.111585.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Copyright on any article is retained by the author(s).
- The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
- The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License
