Publisher

UIN Ar-RaniryUIN Ar-Raniry

Journal Cyberspace: Jurnal Pendidikan Teknologi InformasiCyberspace: Jurnal Pendidikan Teknologi Informasi Abstract

This study aims to compare vegetation classification performance using NDVI derived from Sentinel-2A and Landsat 8 satellite imagery through two different approaches: rule-based classification and machine learning with the Random Forest algorithm. The rule-based approach applies a fixed NDVI threshold of 0.45 to distinguish vegetation and non-vegetation areas. In contrast, the Random Forest model was trained using 70% of the labeled data and tested on the remaining 30%, with NDVI values from both satellite sources as input features. The evaluation results show that the Random Forest model achieved perfect classification accuracy (100%). However, this may be due to using the same labeled dataset for both training and validation, which can lead to overfitting. On the other hand, the rule-based classification yielded an accuracy of 79.7%. This lower performance is likely caused by several factors, including the resolution differences between Sentinel-2 and Landsat 8 imagery, and the subjectivity involved in selecting the NDVI threshold value. The manual threshold setting may lead to bias and a higher number of misclassified pixels. Therefore, while rule-based methods are simple and interpretable, they are less robust. Machine learning approaches, such as Random Forest, offer more flexible and accurate classification when supported by properly separated training and validation datasets.

Conclusion

The study compared rule-based and Random Forest methods for vegetation classification, finding that the Random Forest model achieved perfect accuracy, potentially due to data overlap.The rule-based method yielded lower accuracy (79.7%) due to resolution differences and subjective threshold selection.These results suggest that machine learning approaches are more robust and accurate for vegetation classification when training and validation data are properly separated.

Future Research

Penelitian lanjutan dapat dilakukan dengan memperluas periode waktu analisis dan mengintegrasikan data multisensor (optik dan radar) untuk meningkatkan akurasi klasifikasi vegetasi. Selain itu, eksplorasi algoritma klasifikasi yang lebih canggih, seperti Support Vector Machine (SVM) atau deep learning, dapat dilakukan untuk membandingkan kinerja dengan metode yang telah digunakan dalam penelitian ini. Terakhir, validasi hasil klasifikasi dengan data lapangan yang akurat, seperti survei langsung atau penggunaan citra satelit resolusi tinggi, sangat penting untuk memastikan keandalan dan generalisasi model yang dikembangkan, sehingga dapat diterapkan secara operasional untuk pemantauan vegetasi secara berkelanjutan.

References

1. A fuzzy rule‐based system with decision tree for breast cancer detection - Gupta - 2023 - IET... doi.org/10.1049/ipr2.12774A fuzzy ruleyAAAabased system with decision tree for breast cancer detection Gupta 2023 IET doi 10 1049 ipr2 12774
2. Comparison of Support Vector Machine and Random Forest Methods on Sentinel-2A Imagery for Land Cover... doi.org/10.33022/ijcs.v13i6.4510Comparison of Support Vector Machine and Random Forest Methods on Sentinel 2A Imagery for Land Cover doi 10 33022 ijcs v13i6 4510
3. Known operator learning and hybrid machine learning in medical imaging—a review of the past,... doi.org/10.1088/2516-1091/ac5b13Known operator learning and hybrid machine learning in medical imagingyAAAia review of the past doi 10 1088 2516 1091 ac5b13

Metadata

File size	565.94 KB
Pages	11
DMCA	Report