بررسی عملکرد فتوگرامتری بردکوتاه و پاشش سه بعدی گاوسی در کمّی سازی متغیرهای محیطی در دو حوزه آبخیزداری و جنگلداری

This study examines the utilization of remote sensing techniques for 3D modeling, encompassing traditional close-range photogrammetry and the innovative 3D Gaussian Splatting method. These methodologies are applied to quantify and compute environmental parameters, particularly within the domains of watershed management (erosion) and forestry. Within various areas of Golestan Province, these two techniques have been implemented to address diverse natural resource subjects, yielding findings that demonstrate enhanced accuracy and operational efficiency for both methodologies. Traditional photogrammetry techniques offer superior precision, while contemporary methods deliver accelerated operational speeds while maintaining a harmonious equilibrium between precision and performance. Furthermore, the novel 3D Gaussian Splatting method has successfully surmounted the limitations associated with the field of view in 3D modeling and is adept at modeling expansive environments with heightened speed and visual output. Additionally, certain deficiencies resulting from the photogrammetry method, such as the inability to visually present acceptable details in vegetation cover, can be remedied with the3D Gaussian Splatting method. Therefore, the latter method is more recommended for identifying understory species in forests. However, photogrammetry methods can still be used in research requiring very high centimeter to millimeter precision, such as quantitative volumetric erosion calculations.