Our main line of work in artificial intelligence focuses on the development of advanced point cloud segmentation and classification models.
Unlike generic solutions, we are designing different models tailored to specific scenarios (urban, rural, indoor environments, power lines, etc.), which significantly improves the accuracy and robustness of the results.
The goal is to automate tasks such as:
- Identification of terrain versus objects
- Removal of noise or irrelevant elements
- Extraction of entities (buildings, street lighting, vegetation, etc.)
- Assisted generation of digital terrain models
This approach is already being progressively integrated into tools such as tcp PointCloud Editor, where AI enables more accurate and efficient segmentation of complex scenes.