Over the years, STT has developed its own resources, technologies and methods in machine vision to provide imaging-based automatic checkup and analysis for various applications including automatic inspection, process control and robot guidance in industry.
Some of these methods and resources are:
- Pattern recognition (correlation, edge, shape descriptors, etc.).
- OCR (Optical Character Recognition).
- Classification by SVM (Support Vector Machines).
- Neural networks.
- Stereo vision.
- Colour inspection.
- Blob analysis.
- Edge processing.
- Laser Triangulation.
- Sub-pixel contour extraction.
- MLP (MultiLayer Perceptron) classifiers.
- GMM (Gaussian Mixture Models) classifiers.
Fields of application
- 1D, 2D barcode reading.
- Presence and correct assembly of components (ICs, connectors, etc.).
- All kinds of 2D measurements.
- Reading of registration numbers, sell-by dates, etc.
- Recognition of the position and orientation of objects.
- Guiding of robots.
- Inspection of pad printing-screen printing.
- Filling of containers.
- Detection of scratches or superficial defects.
- 3D reconstruction of components.
- 3D matching.
- 3D measurements.
- Guiding of robots (pick & pack).