Accuracy Assessment of Forest Degradation Detection in Semantic Segmentation based Deep Learning Models with Time-series Satellite Imagery
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
This research aimed to assess the possibility of detecting forest degradation using time-series satellite imagery and three different deep learning-based change detection techniques. The dataset used for the deep learning models was composed of two sets, one based on surface reflectance (SR) spectral information from satellite imagery, combined with Texture Information (GLCM; Gray-Level Co-occurrence Matrix) and terrain information. The deep learning models employed for land cover change detection included image differencing using the Unet semantic segmentation model, multi-encoder Unet model, and multi-encoder Unet++ model. The study found that there was no significant difference in accuracy between the deep learning models for forest degradation detection. Both training and validation accuracies were approximately 89% and 92%, respectively. Among the three deep learning models, the multi-encoder Unet model showed the most efficient analysis time and comparable accuracy. Moreover, models that incorporated both texture and gradient information in addition to spectral information were found to have a higher classification accuracy compared to models that used only spectral information. Overall, the accuracy of forest degradation extraction was outstanding, achieving 98%.
Abstract
This research aimed to assess the possibility of detecting forest degradation using time-series satellite imagery and three different deep learning-based change detection techniques. The dataset used for the deep learning models was composed of two sets, one based on surface reflectance (SR) spectral information from satellite imagery, combined with Texture Information (GLCM; Gray-Level Co-occurrence Matrix) and terrain information. The deep learning models employed for land cover change detection included image differencing using the Unet semantic segmentation model, multi-encoder Unet model, and multi-encoder Unet++ model. The study found that there was no significant difference in accuracy between the deep learning models for forest degradation detection. Both training and validation accuracies were approximately 89% and 92%, respectively. Among the three deep learning models, the multi-encoder Unet model showed the most efficient analysis time and comparable accuracy. Moreover, models that incorporated both texture and gradient information in addition to spectral information were found to have a higher classification accuracy compared to models that used only spectral information. Overall, the accuracy of forest degradation extraction was outstanding, achieving 98%.
참고문헌 (27)
- Cha SE, "Deep Learning-based Forest Fire Classification Evaluation for Application of CAS500-4", Korean J Remote Sens, 38, pp.1273~1283 (2022)
- Clausi DA, "An analysis of co-occurrence texture statistics as a function of grey level quantization", Can J Remote Sens, 28, pp.45~62 (2002)
- Géron A, "Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow" (2022), O’Reilly Media
- Haralick RM, "Textural Features for Image Classification", IEEE Trans Syst Man Cybern, 6, pp.610~621 (1973)
- Huang MH, "Artificial Intelligence in Service", J Serv Res, 21, pp.155~172 (2018)
- Kim ES, "Forest Management Research using Optical Sensors and Remote Sensing Technologies", Korean J Remote Sens, 35, pp.1031~1035 (2019)
- "K-Forest" (2024)[URL]
- Kross A, "Assessment of RapidEye vegetation indices for estimation of leaf area index and biomass in corn and soybean crops", Int J Appl Earth Observ Geoinf, 34, pp.235~248 (2015)
- Lee WK, "Application of Remote Sensing and Geographic Information System in Forest Sector", J Cadastre Land InformatiX, 46, pp.27~42 (2016)
- Li W, "Urban Water Extraction with UAV High-Resolution Remote Sensing Data Based on an Improved U-Net Model", Remote Sens, 13, pp.3165~ (2021)
- "Fixing Weight Decay Regularization in Adam" (2018)[URL]
- "Cadastral Statistics" (2023)[URL]
- Park EB, "Comparison of Sampling and Wall-to-Wall Methodologies for Reporting the GHG Inventory of the LULUCF Sector in Korea", J Clim Chang Res, 9, pp.385~398 (2018)
- Ronneberger O, "U-Net: Convolutional Networks for Biomedical Image Segmentation", Medical Image Computing and Computer-Assisted Intervention -MICCAI 2015, pp.234~241 (2015), Springer
- Rouhi R, "Benign and malignant breast tumors classification based on region growing and CNN segmentation", Expert Syst Appl, 42, pp.990~1002 (2015)
- Rumelhart DE, "Learning representations by back-propagating errors", Nature, 323, pp.533~536 (1986)
- Smith LN, "Super-convergence: very fast training of neural networks using large learning rates", SPIE, 11006, pp.369~386 (2019)
- Solórzano JV, "Land Use Land Cover Classification with U-Net : Advantages of Combining Sentinel-1 and Sentinel-2 Imagery", Remote Sens, 13, pp.3600~ (2021)
- Son SH, "Land-Cover-Change Detection with Aerial Orthoimagery Using SegNet-Based Semantic Segmentation in Namyangju City, South Korea", Sustainability, 14, pp.12321~ (2022)
- Woo HS, "Precision Forestry Using Remote Sensing Techniques : Opportunities and Limitations of Remote Sensing Application in Forestry", Korean J Remote Sens, 35, pp.1067~1082 (2019)
- Zhang P, "Urban Land Use and Land Cover Classification Using Novel Deep Learning Models Based on High Spatial Resolution Satellite Imagery", Sensors, 18, pp.3717~ (2018)
- Zheng X, "Semantic Segmentation Model for Wide-Area Coseismic Landslide Extraction Based on Embedded Multichannel Spectral–Topographic Feature Fusion : A Case Study of the Jiuzhaigou Ms7. 0 Earthquake in Sichuan, China", Remote Sens, 15, pp.1084~ (2023)
- Zhou Z, "Deep learning in medical image analysis and multimodal learning for clinical decision support", pp.3~11 (2018), Springer