Photorealistic 3D models digitally represent the real world, showing real forms of buildings, bridges, landforms, etc., with up-to-dateness. Using multi-source data (images, videos, point clouds, thermal infrared), software processes—feature extraction, homologous matching, meshing, texturing—generate texture-realistic models that measure object size and volume. They support fusion of multi-source, indoor-outdoor, and air-ground-space data, serving as a digital twin base for industries.

Digital Elevation Model (DEM) digitally simulates terrain with limited elevation data, representing elevation via ordered numerical arrays. It generates contours, perspectives, etc., supporting spatial composites, surface/line analysis.Digital Orthophoto Map (DOM), using DEM, corrects digitized aerial photos pixel-by-pixel, then mosaicked and frame-cropped. It combines map geometric accuracy and image features, with real textures for map data, used in mapping, land surveying, planning & design, etc.They both form composite landscape maps, showing terrain undulations and ground textures.

Point cloud data carries rich 3D structure, effectively bridging 2D images' lack of spatial structure in applications. Properties: coordinates, normal vectors, color, object reflective intensity. Raw data includes noise, outliers, holes, redundancy, requiring processing (initialization, registration, fusion, denoising). Classified by pose, time, spatial topology, it extracts buildings, roads, boundaries, barriers, roadside trees, poles, traffic signs—used in intelligent navigation, smart cities, cultural heritage preservation, etc.