Computer Vision, Object Pose Estimation, and Process Insights

Quality inspection using machine vision and deep learning within manufacturing is a growing industry. Accurately identifying quality defects using machine vision is the initial step in a comprehensive defect prevention solution. Eigen Innovations has created internal toolsets for significantly improving the performance of deep learning models through dataset standardization approaches that leverage CAD, inspection images, and specialized algorithms. Several significant enhancements are being developed through this project to enable the productization of these tools. Once these defects can be identified, preventing them from occurring in the process is the next step in a defect prevention solution. A second aspect of this project involves leveraging process outcomes obtained with deep learning applied to machine vision data to identify causal relationships with the process data and control parameters. This will allow Eigen to identify process contexts and provide operators with actionable recommendations for process improvements.

Faculty Supervisor:

Rickey Dubay

Student:

Partner:

Eigen Innovations Inc

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

University of New Brunswick

Program: