SimActive Enhances Correlator3D with Gaussian Splatting and Photogrammetry Integration

Gaussian Splatting in 3D Reconstruction

Gaussian splatting has emerged as a significant development in the field of 3D reconstruction, known for its ability to create realistic and smooth-rendered scenes. This advancement raises a critical question within the geospatial industry: can it extend beyond visualization to support practical mapping workflows?

SimActive has addressed this question with the release of Correlator3D Version 11, which integrates Gaussian splatting directly into its photogrammetry platform.

Enhanced Visual Capabilities

Gaussian splatting utilizes dense collections of Gaussian primitives to represent 3D environments, resulting in continuous surfaces that often appear more natural than traditional mesh-based models. This technique is particularly effective in complex environments where texture and depth perception are crucial.

In Correlator3D Version 11, users can now generate splat-based models directly from imagery and utilize them alongside standard photogrammetry outputs. According to SimActive, the integration aims to deliver high-quality 3D visualization while preserving established production workflows. CTO Louis Simard stated, “Version 11 gives users a practical way to produce high-quality 3D content with stunning visual fidelity while maintaining full control within a familiar photogrammetric environment.”

Distinction Between Visualization and Measurement

The incorporation of Gaussian splatting into professional software emphasizes an important distinction: visual quality does not equate to measurable accuracy. Gaussian splatting is adept at creating immersive scenes that are intuitive and easy to interpret, even when the underlying data may be inconsistent or incomplete.

In contrast, photogrammetry is designed for precision, providing georeferenced datasets and consistent geometry that support engineering, surveying, and GIS applications. These characteristics are essential for workflows that rely on reliable measurements. Rather than replacing these outputs, SimActive positions Gaussian splatting as a complementary layer.

Combining Approaches in a Unified Workflow

This release is notable not only for introducing a new visualization method but also for enabling the combination of this method with established photogrammetry processes within a single environment. Users can transition from raw imagery to both accurate geospatial products and visually rich 3D models without needing to switch platforms. This integration reflects a broader industry trend where visualization is increasingly seen as an integral part of data delivery.

Correlator3D Version 11 allows users to:

• Generate Gaussian splat models directly from image datasets
• Visualize scenes with enhanced realism and smooth rendering
• Edit and refine splat outputs within the same software environment
• Maintain access to traditional photogrammetry products for measurement and analysis

This unified approach enables teams to customize outputs for various audiences, including engineers, analysts, and stakeholders who require clear visual context.

The Role of Gaussian Splatting Today

Gaussian splatting serves as an enhancement to visualization rather than a substitute for measurement workflows. Its advantages are particularly evident in applications where realism and clarity facilitate understanding, such as inspection, planning, and communication. Meanwhile, traditional photogrammetry continues to provide the accuracy and consistency necessary for operational decision-making.

By integrating both approaches, SimActive allows users to leverage the strengths of each without compromise. Accurate data remains the foundation, while Gaussian splatting offers a new perspective on how that data can be interpreted and presented.

A Step Forward in Geospatial Technology

The integration of Gaussian splatting into Correlator3D signifies a practical advancement for the technology, transitioning the discussion from experimentation to application. This development illustrates how new visualization methods can be incorporated into established geospatial workflows.

Currently, the roles are clearly defined: photogrammetry provides the measurable foundation, while Gaussian splatting enhances the experience of that information. Together, they suggest a future where accuracy and realism coexist as integral components of the same workflow.