WolkenRefinery
Point cloud segmentation and annotation software
WolkenRefinery is a point cloud classification correction tool built on the CloudCompare point cloud editing platform. It imports semantically labeled point clouds (e.g., outputs from WolkenInsight) and uses multi-dimensional information — such as classification confidence scores and RGB color data — to detect misclassified points. WolkenRefinery then allows efficient batch correction of these points, updating them to their proper classes.
Why WolkenRefinery
Revolutionary Modeling Efficiency
Uses AI to identify complex shapes in point clouds, boosting modeling speed by up to 90%.
Millimeter-level Modeling Accuracy
Its deep learning models improve boundary recognition, using WolkenInsight’s semantic labels to distinguish beams, columns, and pipelines with millimeter-level precision.
Component-level Semantic Modeling
Automatically identifies and reconstructs key building components, enabling extraction and optimization of complex contours.
City-level Modeling
Capability
Supports rapid city-scale point cloud modeling by using parallel processing (multi-task concurrency) to execute batch projects efficiently.
Industry-Specific
Adaptation
Offers built-in support for 20+ building code modeling rules, covering both national and European standards.
Strong Platform
Compatibility
Seamlessly integrates with Rhino and other popular modeling platforms, with support for a wide range of point cloud and CAD/BIM format conversions.
Functions
Semantic Classification Correction
Multidimensional Data Fusion
Simple Correction Process
Semantic Classification Correction
Combine AI confidence levels to identify misclassifications and enable efficient manual correction.
Multidimensional Data Fusion
Integrate RGB color, spatial structure, confidence scores, and other information to locate anomalies.
Simple Correction Process
The process is easy to operate, saving time with large-scale batch processing.
Applications
With its powerful features, WolkenRefinery has extensive and important application scenarios in many fields

Data Refinement
(BIM Context)

Digital Precision Enhancement
(Cultural Heritage Context)

Infrastructure Classification
(Transportation Context)

Reconstruction Optimization
(Industrial Context)

Outputting High-Precision Classified Data
By integrating RGB color data with AI confidence metrics, it refines the segmentation results and outputs highly accurate, classified point clouds. This provides more reliable geometric and semantic baseline data, enabling downstream BIM modeling to be built on a solid, precise foundation.
Enhancing BIM Semantic Foundations
In Building Information Modeling (BIM) workflows, point cloud data often undergoes AI semantic segmentation followed by labor-intensive manual checks. WolkenRefinery revolutionizes this process by swiftly locating semantic misclassifications — such as columns mistakenly labeled as walls, or missing doors and windows.
Enabling Accurate Heritage Preservation
By guiding targeted manual corrections, it elevates overall semantic accuracy. This strengthens digital preservation and facilitates the dissemination of cultural heritage assets, ensuring that their historical and artistic value is authentically retained in digital form.
Securing Authentic Digital Restoration
In cultural heritage digital modeling, precise semantic classification is essential for authentic visualization, restoration, and archival. WolkenRefinery detects subtle misclassifications caused by intricate textures or structural damage, such as eaves mistaken for walls or carvings blended into background surfaces.
Supporting Safer & Smarter Infrastructure
The refined semantic outputs provide trustworthy data for road safety diagnostics, structural health monitoring, and optimized operations. This supports more proactive maintenance planning and enhances the resilience of transportation infrastructure.
Improving Transportation Facility Insights
In semantic processing of point clouds from roads, bridges, and tunnels, mislabeling of edges or cross-category confusion often hampers maintenance and safety assessments. WolkenRefinery leverages confidence-based tools to identify and rectify such issues, ensuring features like railings, slopes, and tunnel linings are correctly classified.
Establishing Closed-Loop Smart Refinement
The corrected outputs can also serve as training samples for iterative model updates, building an industry-specific closed-loop AI segmentation system. This enhances the precision and efficiency of industrial digital transformation initiatives.
Empowering Precise Industrial Modeling
In industrial plants and warehouses, automated segmentation can sometimes confuse machinery with structural components. WolkenRefinery enables semantic correction of key assets like pipes, valves, and frameworks, ensuring critical equipment is distinctly identified within digital reconstructions.

Learn More

Contact Us

Share your vision with us
— we’ll bring the technology to make it real.

Thank you for reaching out!
Our team will be in touch shortly.
Tell us about your project so we can tailor the best solution for you.

Fill Out Now

© 2025 WolkenVision AG.