Wood Species Recognition ALGORITHM
A YOLO-based computer vision model trained to classify different wood species in real time. Integrated with a webcam and robotic fabrication workflow, the system enables:
Automated species detection during milling.
Adaptive toolpath and speed adjustments based on material.
Potential applications in timber sorting and quality control.
Next steps include expanding the dataset and refining accuracy under varied lighting and surface conditions.
Institution: University of Pennsylvania
Lab: Advanced Research and Innovation lab
Instructor: Patrick Danahy, Alicia Nahmad Vazquez
Co-Instructor: Mahsa Masalegoo
Lab Managers: Nicholas Sideropoulos, Shunta Moriuchi
In Collaboration with: Burcu Gocen, Zitong Ren, Qingyang Xu.
Agentive Ai System
Automated recognition of wood species with adaptive robotic responses.
the project introduces machine learning-driven fabrication, utilizing both a YOLO-based wood detection model and a sound classification system to autonomously adjust robotic milling speeds based on wood species.
