1 August 2018 in TELKOMNIKA (Telecommunication Computing Electronics and Control)
The density level in the leaf venation type has different characteristics. These different characteristics explain the environment in which plants grow, such as habitat, vegetation, physiology and climate. This research aims to measure of leaf venation density, leaf venation feature analysis and then identifying plants based on venation type. Stages of this research include leaf image data collection, segmentation, vein detection, feature extraction, feature selection, classification, evaluation and ending with analysis. The results of this study indicate that the level of leaf venation density is quite good is the type of venation paralellodromous, acrodromous and pinnate. Based on the selection of features using Boruta Algorithm, obtained 19 most important features that represent the type of leaf venation. This is reinforced by the average of accuracy produced at the time of classification using SVM, which amounted to 77.57%.
1 January 2017 in IOP Conference Series: Earth and Environmental Science
The forest fires in Indonesia occurs frequently in the dry season. Almost all the causes of forest fires are caused by the human activity itself. The impact of forest fires is the loss of biodiversity, pollution hazard and harm the economy of surrounding communities. To prevent fires required the method, one of them with spatial temporal clustering. Spatial temporal clustering formed grouping data so that the results of these groupings can be used as initial information on fire prevention. To analyze the fires, used hotspot data as early indicator of fire spot. Hotspot data consists of spatial and temporal dimensions can be processed using the Spatial Temporal Clustering with Kulldorff Scan Statistic (KSS). The result of this research is to the effectiveness of KSS method to cluster spatial hotspot in a case within Riau Province and produces two types of clusters, most cluster and secondary cluster. This cluster can be used as an early fire warning information.
10 October 2025 in IEEE Access
Accurate tree crown delineation from very high-resolution UAV orthophotos in urban areas remains challenging due to landscape complexity and the scarcity of representative annotated data. This study proposes an enhanced Mask R-CNN approach, supported by semi-automatic annotation using the Segment Anything Model-HQ (ViT-H Quant), to improve tree crown delineation. The approach evaluates the effect of backbone architectures (ResNet50 and ResNet101) and transfer learning strategies from general (COCO) and domain-specific datasets (OAM-TCD and Detectree2). Four model variants were tested through 5-fold cross-validation and independent test data evaluation. Results demonstrate that the model with a ResNet50 backbone pre-trained on OAM-TCD achieved the best performance (AP 49.56%, F1-score 0.722, precision 0.720, recall 0.723), while also being computationally more efficient than ResNet101. Although Wilcoxon signed-rank tests indicated that the improvements were not statistically significant (p > 0.05), consistent performance trends still favored the OAM-TCD–based model. These findings highlight the importance of domain-specific pretraining and computational efficiency for practical large-scale UAV-based urban vegetation monitoring.
1 August 2023 in Indonesian Journal of Artificial Intelligence and Data Mining
Intertidal mangrove forests are ecosystems that are extremely productive offering diverse socio-economic advantages. Preserving and appropriately using these ecosystems is crucial. However, safeguarding and restoring mangroves present challenges due to their extensive and hard-to-reach areas. Leveraging remote sensing technology and diverse image classification methods has shown promise in accurately mapping and monitoring mangroves. This study reviews the use of machine learning methods in mapping and monitoring mangroves, particularly using moderate-resolution multispectral satellite images. The literature study was conducted by systematically searching and analyzing articles published in Scopus-indexed journals from 2018 and 2023. The primary goals are to uncover methodologies for mapping mangroves with moderate-resolution imagery, identify advancements in machine learning algorithms, and assist researchers in staying updated in this field. The findings reveal that various machine-learning algorithms can be employed to map mangroves. Mangrove mapping with machine learning typically involves stages such as inputting multispectral images, image preprocessing, image classification, and assessing accuracy. Among the techniques, in the case of remote sensing data, ensemble tree-based approaches such as random forest outperform single classifiers. Potential and emerging issues for future research encompass automating the generation of training datasets for specific land cover classification, developing methods to transfer the classification model to different study areas, and making use of cloud-based technologies for processing remote sensing data.