Waterloo, Canada – August 10, 2021

Teledyne is pleased to announce the release of Sapera Vision Software Edition 2021-07. This powerful tool from Teledyne DALSA delivers proven image acquisition, control, processing and artificial intelligence capabilities for building high-performance machine vision applications.

Key enhancements in this update include significant improvements to Astrocyte™, the AI training graphical tool, and the Sapera Processing environment. These upgrades expand the software’s ability to deliver robust image analysis solutions.

“We’re excited about our new continual learning feature,” said Bruno Ménard, Director of Software for Teledyne’s vision solutions group. “This innovation allows users to improve classifier accuracy in the field by focusing only on misclassified samples—eliminating the need for complete retraining.” The update also introduces pixel-level anomaly detection with runtime heatmap generation capabilities.

Sapera Vision excels in applications like metal surface inspection, hardware part identification, vehicle detection and segmentation, and x-ray noise reduction. These upgrades enhance its ability to deliver precise results across industrial, medical, and research settings.

Notable Improvements:

  • Continual Classification: Users can now pre-train classifiers in Astrocyte and continue refining models during runtime processing
  • Enhanced Anomaly Detection: New algorithm provides more accurate defect localization with heatmap output capabilities
  • Simplified Dataset Creation: Live video acquisition enables automatic dataset generation for training purposes

For additional details, please visit our website.

About Teledyne With decades of combined expertise across imaging technologies, Teledyne offers comprehensive solutions spanning aerospace, industrial inspection, scientific research, medical technology, and more. Their integrated approach delivers specialized tools backed by global technical support to tackle even the most challenging imaging requirements worldwide.

Last Updated: 2025-09-05 04:55:57