Your Guide to CHC Navigation's Integrated 3DGS Solution

A complete 3D Gaussian Splatting (3DGS) workflow involves more than capturing data. Hardware integration, data processing and model visualization all play a part in turning field data into usable 3D results. This guide answers the questions users ask most often about the integrated 3DGS solution from CHC Navigation, in the order the work is done: how to pair the CHCNAV RS7 handheld scanner with an Insta360 panoramic camera, how to process the data in the cloud, and how to view and share the finished models. If you are new to the technology itself, our article on mesh modeling and 3D Gaussian Splatting explains how 3DGS works.

The full workflow is shown in our video: RS7 + Insta360 X6, 3D Gaussian Splatting workflow.

Data Capture

Which Insta360 cameras work with the RS7 and the RS10?

Full Insta360 X6 support for the RS10 is planned for a future version. To see the RS7 and an Insta360 X6 set up and processed step by step, watch the RS7 + Insta360 X6 tutorial on LiDAR scanning and CoPre processing.

The RS7 already has cameras. What does an Insta360 add?

With its built-in cameras, the RS7 captures everything in a single scanning session and produces the essential deliverables: colored point clouds, meshes and 3DGS models. Its dual 12 MP HD cameras and the HPGS 2.0 engine are designed for real-color scene reconstruction, which covers the modeling needs of routine work.
 

Pairing the RS7 with an Insta360 panoramic camera is an upgrade for professional-grade applications. Insta360 has mature panoramic imaging technology, with strong image quality in difficult conditions such as high dynamic range scenes and low light. Combined with the spatial data from the RS7, it enhances the texture detail and visual quality of 3DGS models, which helps most in low light and in interiors with complex lighting.

Compared with using a panoramic camera on its own, combining LiDAR point clouds with panoramic imagery has three advantages.
 

More accurate real-world scale. Purely visual modeling can struggle to reproduce real dimensions. LiDAR fusion improves the scale accuracy of the model and makes it more reliable for engineering measurements.

More stable reconstruction where there is little texture. In areas with limited visual texture, such as plain walls and open spaces, purely visual solutions can fail to match features, drift over a large capture area and produce misaligned models. Sensor fusion keeps the camera position estimate stable in these low-feature environments.

Several 3D deliverables from one capture, with consistent quality. With LiDAR fusion, the same data produces not only a 3DGS model but also point clouds, meshes and other 3D deliverables, so one capture serves several uses.

What does the solution improve in large, dark and crowded spaces?

This upgrade focuses on reconstruction in difficult environments and addresses three common problems on site:
 

• Large areas. On large projects and in spaces with complex structures, the solution helps avoid gaps, misalignment and missing sections, so the panoramic reconstruction stays stable across the whole capture.
   
• Low light and complex indoor lighting. The imaging capability of the panoramic camera, together with algorithm optimization, improves texture clarity in dark interiors and in scenes with strong contrast between light and shade, with richer detail and less visible noise.
   
• Crowded scenes. An AI-based dynamic pedestrian filtering algorithm detects and removes people who move through the scene, which reduces ghosting and gives a cleaner model of the actual space.

How long should one capture session be, and what computer is needed?

• Capture time. Keep each session with the RS7 and the Insta360 X6 under 30 minutes. The longer the session, the less efficient the later processing becomes.
   
• Computer for desktop processing. 64 GB of RAM and an NVIDIA RTX 4090 graphics card with 24 GB of video memory are recommended. Processing in CoCloud, described below, removes this requirement.

Data Processing

Why is cloud processing with CoCloud recommended?

CoCloud removes the need for a high-performance computer. Upload the field data and the 3DGS model is generated automatically in the cloud. Compared with desktop software, there is no software to install and no complex parameters to set, and processing is faster.
 

Before processing starts, the user chooses the outputs. Alongside the 3DGS model, listed as HPGS, CoCloud can produce real-time and optimized point clouds (LAS), panoramic images, the original images (JPG), an orthophoto (TIF) and a mesh model (OSJB and OBJ), and it shows the estimated processing time for each before the job begins.

Computing resources in the cloud scale to the size of the job. Once processing is complete, the results can be viewed, managed and shared online, which keeps the whole 3DGS production in one place. You can sign in to CoCloud directly, and the CoCloud product page covers measurement, annotation, sharing and deployment options.

Data Visualization

What is CoViewer, and what makes it different?

CoViewer is a free, lightweight desktop 3DGS viewer developed by CHC Navigation. It gives smooth interaction and high-definition detail, which makes it suited to presenting 3D deliverables, particularly from handheld scanning and complex indoor environments. Our CoViewer launch article covers its file formats and navigation modes. Here the focus is the rendering engine behind it.
 

CoViewer runs on the UniRender engine developed by CHC Navigation, which optimizes the whole 3DGS rendering pipeline with four technologies:

Typical uses include painted details in cultural heritage sites and exhibition halls, wood grain and fabric in residential interiors, and complex interiors with strong contrast between light and shade. Detail and color hold up even when viewed close.

How do CoCloud and CoViewer work together?

From capture to cloud processing to desktop viewing, the workflow runs in four steps:
 

1. Capture with the integrated 3DGS solution

RS series handheld scanner with an Insta360 camera
 

2. Upload the data to CoCloud for processing

Local computing and hardware requirements drop sharply
 

3. Download the deliverables to CoViewer in one click

High-definition offline viewing, without unstable site networks, browser lag or image compression
 

4. Present the results

Presentations, design reviews, digital cultural tourism and digital twin displays, with high-definition visuals and smooth interaction

Which file formats are supported, and how is HSC different from PLY?

Two output formats are supported, one for general use and one for professional, high-precision work:
 

• PLY. The industry-standard format, compatible with common 3D viewing and editing tools.
   
• HSC. The CHC Navigation format, with three advantages over PLY.

One Workflow from Capture to Presentation

The integrated 3DGS solution from CHC Navigation connects data capture, cloud processing and 3D visualization in one workflow. It simplifies the field work, improves how results are viewed and shortens the time from site visit to finished model. For a field example with the same scanner, see our indoor survey workflow with the RS7, and for the full range of handheld, mobile and airborne systems, our 3D mobile mapping solutions.

____

About CHC Navigation

CHC Navigation (CHCNAV) develops advanced mapping, navigation, and positioning solutions designed to increase productivity and efficiency. Serving industries such as geospatial, agriculture, machine control and autonomy, CHCNAV delivers innovative technologies that empower professionals and drive industry advancement. With a global presence spanning over 140 countries and a team of more than 2,200 professionals, CHC Navigation is recognized as a leader in the geospatial industry and beyond. For more information about CHC Navigation [300627.SZ], please visit: https://geospatial.chcnav.com/about/overview