- ES
- |
Point Cloud Processing Engine
An internal tool for point cloud processing used across various projects via pipelines with a cloud-based serverless architecture.
Designed for maximum versatility and ease of use, this tool allows other projects to delegate high-computational-cost point cloud processing without deploying a dedicated infrastructure. Pipeline-based task definition, composed of generic processing steps, enhances modularity and facilitates code reuse across projects.
Point Clouds
Point clouds are an essential tool for capturing and analyzing reality with high precision, allowing for the detailed digitization of physical environments, facilitating inspection, modeling, simulation, and terrain analysis. The challenge is not just data acquisition but efficient processing: handling millions of points, filtering noise, segmenting structures, and extracting relevant information requires specialized and high-performance tools.
This cloud-based processing solution offers high scalability, parallel computing, faster execution times, reduced development workload for new projects, and easier future maintenance.
Pipelines
- Processing tasks are defined in a simple JSON syntax as multi-step pipelines, offering maximum flexibility. There are two types of processing steps:
- PDAL Stages: fast, efficient, and validated by a large open-source developer community.
- Custom Python Functions: developed for this tool when PDAL falls short and specific calculations are required.
Multi-Project
- Its modular syntax, parallel processing, and scalable architecture enable simultaneous use by multiple client tools requiring point cloud processing.
Multi-Input
- Designed for usability and adaptability, data can be input via API, Lambda functions, S3 automations, or MQ messages to suit client applications.
Serverless Cloud Processing
- All processing tasks run entirely in a cloud-based serverless environment, ensuring engine availability and facilitating future scalability.