| تعداد نشریات | 45 |
| تعداد شمارهها | 1,562 |
| تعداد مقالات | 19,103 |
| تعداد مشاهده مقاله | 65,993,745 |
| تعداد دریافت فایل اصل مقاله | 24,025,144 |
Automated Detection and Mapping of Date Palm Trees in Urban Green Spaces Using Deep Learning and Google Earth Imagery | ||
| نشریه کاربرد سنجش از دور و سیستم اطلاعات جغرافیایی در علوم محیطی | ||
| دوره 6، شماره 21، دی 1405 | ||
| نوع مقاله: مقاله پژوهشی | ||
| شناسه دیجیتال (DOI): 10.22034/rsgi.2026.73748.1170 | ||
| نویسندگان | ||
| محمد عباسی* 1؛ مصطفی کابلی زاده2 | ||
| 1استادیار گروه نقشهبرداری، دانشکده فنی و مهندسی، موسسه آموزش عالی عمران و توسعه، همدان، ایران | ||
| 2دانشیار گروه سنجش از دور و سیستم اطلاعات جغرافیایی، دانشکده علوم زمین، دانشگاه شهید چمران، اهواز، ایران | ||
| چکیده | ||
| Identifying the spatial location of trees is the first step towards managing green spaces and analyzing the spatial distribution of trees in urban areas. In this context, the date palm holds particular significance in the southern regions of Iran. Generating tree location maps through ground-based surveying methods requires substantial time and financial resources. Hence, the use of aerial or satellite imagery is recommended. This study utilizes high-resolution satellite images downloaded from Google Earth to detect and extract date palm trees. Deep learning methods including RetinaNet, Faster R-CNN (FRCNN), and Single Shot Multibox Detector (SSD) were employed. Initially, over 2,000 image samples were cropped into two classes: containing date palms and background, with dimensions of 128 x 128 pixels. The models were trained on 80% of the samples for learning and validated on the remaining 20% over 20 training epochs. Subsequently, the trained models were applied to test images. The results indicate that the RetinaNet model achieved a precision of 0.90, a recall of 0.94, and an F1 score of 0.92, demonstrating optimal processing time. By automatically extracting date palm trees from Google Earth satellite images, a spatial database of date palm locations was created. Finally, maps displaying the spatial distribution of date palms were developed for optimal management of urban green spaces within a GIS environment. The findings suggest that it is feasible to generate a distribution map of date palm locations from Google Earth satellite images using deep learning methods at an appropriate cost. | ||
| کلیدواژهها | ||
| Deep learning؛ Palm tree؛ Google Earth؛ Satellite images | ||
| مراجع | ||
|
| ||
|
آمار تعداد مشاهده مقاله: 35 |
||