Urban Modeling: LiDAR + Imagery Mapping of Maysville, Kentucky
Maysville, Kentucky is a small Ohio River town with roots that go back to the late 1700s. The town grew as a riverport built on tobacco trade, early industry, and steady commercial traffic along the Ohio River. The historic center developed around narrow streets, brick storefronts, and steep riverbank terrain, creating a dense and varied built environment that is difficult to map with traditional methods. These characteristics make Maysville an ideal location to demonstrate the strengths of a LiDAR and imagery workflow.
Mapping an urban environment is never simple. You will run into narrow streets, reflective metal roofs, and many diverse vertical facades that can create blind spots and distortions when using traditional photogrammetry. For surveyors and city planners, those gaps make it difficult to produce the complete, measurable models needed for accurate planning and design.
By using a LiDAR + imagery workflow, users are able to reap the benefits of getting a very accurate LiDAR registration point cloud, along with the texture and color point cloud that is produced by an RGB camera. Pairing the two together provides both geometric precision and visual context. Ultimately, a dataset that captures true building facades, pavement elevations, and urban features with minimal occlusions. Within PixElement, these datasets can be viewed, classified, and merged into a single, cohesive model that supports linework extraction, surface generation, and detailed visualization.
This approach was put into action by Aerial Image Solutions to create a model of the Maysville, Kentucky using PixElement. Their goal was to create a high-resolution 3D model that captured every curb, rooftop, and street feature with accuracy.
By combining LiDAR and RGB imagery within PixElement, the team captured both the structural precision of LiDAR and the visual richness of high-resolution imagery. This hybrid workflow provided a comprehensive view of Maysville’s urban landscape.
Orthoimage vs DEM - Drag the Slider to Compare
Urban environments present unique challenges for surface reconstruction: relative metal roofs, parked or moving vehicles, and complex rooflines can cause more noise or gaps in traditional photogrammetry. LiDAR overcomes these obstacles by directly capturing elevation and structure through laser returns, while imagery provides the detailed texture and color needed for visualization and communication. By incorporating LiDAR, a denser and more complete point cloud is achieved, improving geometric precision while maintaining color fidelity through co-registered RGB imagery.
Right: LiDAR point cloud visualized with RGB color mapped from co-registered imagery.
Within PixElement, these datasets were processed, registered, and visualized side by side, enabling point cloud refinement, surface reconstruction, and 3D visualization in one connected environment. The result is a detailed digital model of Maysville that merges the strengths of both capture methods into a single, unified dataset.
LiDAR and Photogrammetry each hold a specific set of tools and abilities that should be utilized in certain situations. For example, you would use LiDAR if you wanted to capture very intricate details or thin surfaces such as narrow building surfaces (such as church or courthouse steeples), a powerline study, or model the terrain under a forest canopy. You would use photogrammetry if you wanted to capture high quality textures and rich colors, such as creating a model of a historic building, construction site, or quarry where surface detail, visual context, and color accuracy matter. LiDAR provides the geometric skeleton with precise elevation data, while photogrammetry delivers the rich realistic textures.
PixElement bridges the gap of LiDAR and Photogrammetry’s individual features and merges the best of both technologies. PixElement gives users the flexibility to choose the right approach for their project.
| Feature | LiDAR | Photogrammetry |
|---|---|---|
| Accuracy | Very high geometric accuracy (independent of lighting) | High accuracy with proper GCPs or RTK/PPK integration |
| Vegetation Penetration | Yes, can capture terrain beneath canopy | No, limited to visible surfaces |
| Light Dependence | Works day or night. No lighting restrictions | Requires good lighting conditions. Highly influenced by lighting, camera quality, and lack of precise distance measurements |
| Cost | Higher equipment and operational cost | Lower cost, more accessible |
| Color Information | Minimal visual detail | Rich color and surface texture |
| Ease of Use | Requires specialized setup and calibration | Easier to deploy with standard drones |
| Versatility | Excellent for terrain/topographical surveys, infrastructure, and forestry | Ideal for construction, inspection, and visualization |
| Processing Time | Fast acquisition, moderate processing | Slower capture, more compute-heavy reconstruction |
In this project, the LiDAR + imagery workflow produced a noticeably cleaner and more complete reconstruction of Maysville’s built environment. Building facades, curb lines, and vertical structures remained consistent across the dataset with fewer gaps and/or distortions than an imagery-only workflow.
The RGB-only reconstruction—using the same tiepoints and flight images—showed missing geometry on metal roofs, thin architectural elements, and areas with strong shadows or occlusions. The integrated LiDAR dataset preserved these details, improving measurement reliability and providing survey-grade structure while still maintaining high-quality color and texture.
Building Facades & Narrow Architecture
Features like church steeples, parapet walls, and multi-story facades expose the limits of image-only reconstruction. These structures rely on sharp vertical geometry and thin architectural elements that photogrammetry often softens or loses entirely, especially when shadows or narrow viewing angles are involved. In the LiDAR+Imagery model, these same features remain intact: straight walls stay straight, rooflines hold their form, and slender details are captured as measurable geometry.
Narrow and Shadowed Streets/Alleyways
Narrow streets and shadowed alleyways introduce some of the most persistent reconstruction gaps in urban photogrammetry. Limited camera angles, deep shadows, and tight building spacing reduce feature visibility, causing RGB-only models to lose wall geometry, merge surfaces, or leave voids altogether. In the LiDAR-supported reconstruction, ground surfaces remain defined, and occluded areas are filled win with accurate elevation data.
Reflective Roofs
Reflective and low-texture roof surfaces, common on metal buildings across Maysville, are a consistent weak point for imagery-only reconstruction. Photogrammetry relies on stable visual features to generate depth, and bright metal roofs offer almost none. Highlights shift between images, shadows move, and large uniform areas provide no texture to match. The result is missing roof sections, distorted ridgelines, and irregular surface patches. The LiDAR-enhanced workflow mitigates these issues by contributing direct elevation measurements, even when the roof reflects light or lacks visual detail. While LiDAR can produce fewer returns on highly reflective surfaces, the geometry it does capture remains structurally accurate, allowing the final model to preserve roof shape, pitch, and edge definition far better than RGB alone.
Delivering Reliable Urban Models Through Hybrid Capture
The Maysville dataset shows why LiDAR and imagery are no longer competing approaches, they are complementary tools that solve different pieces of the semi-urban/urban-mapping puzzle. LiDAR delivers the structural truth needed for confident measurement, while imagery provides the clarity and context that support communication, planning, and design. By fusing the two with PixElement, surveyors and municipalities gain a model that is both geometrically dependable and visually intuitive.
References:
Wu, Bo. “Photogrammetry for 3D Mapping in Urban Areas.” Urban Informatics, edited by Wenhao Shi, Michael F. Goodchild, Michael Batty, Mei-Po Kwan, and Anthony Zhang, The Urban Book Series, Springer, 2021, doi.org/10.1007/978-981-15-8983-6_23 .
Guo, Liang, et al. “Extraction of Dense Urban Buildings from Photogrammetric and LiDAR Point Clouds.” IEEE Access, vol. 9, 2021, pp. 111823–111832, doi.org/10.1109/ACCESS.2021.3102632 .
All images sourced from PixElement.