Image based decisions require broad competence in technologies from high speed computing and image processing, to machine learning and pattern recognition. DRVISION's patented image based decision technologies are comprehensive and have proven effectiveness.
High Speed Processing and Tracking
Time-lapse microscopy sequences have created one of the more challenging applications in the image inspection field – cell and particle tracking. DRVISION has best-in-class tracking algorithms developed with world class collaborators. Our processing methods are orders of magnitude faster than conventional methods.
DRVISION has developed special high speed and high precision search algorithms. Our methods achieve significantly higher search speed (up to 5,000x) and robustness over other technology.
Structure-Guided Processing technology enhances image features and accurately detects structures of interest in noisy and low contrast conditions, and can compensate for image variations and distortions with high accuracy, repeatability and accuracy. A general purpose technology, it has been very useful for live cell microscopy experiments where the image characteristics (e.g. noise and signal level, optical distortions) change over time.
Dynamic Decision and Control
DRVISION has fundamental technologies on object based boundary refinement, robust high content informatics, adaptive image region partition and morphologic processing. DRVision technology includes a versatile Recognition Frame for unified image processing and data visualization.
DRVISION's learning technologies enable intelligent systems for broad applications. Application engineers can create high performance solutions through fast and user friendly teaching interface. Learning technologies and patents cover segmentation learning, partition learning, decision learning and recipe creation.
Through a collaborative partnership with Nikon, DRVISION has developed a suite of robust methods and models for intelligent recognition for image-based decision. These decision technologies provide the detection of structures of interest in images or image sequences, quantitiative characterization of those strucutres, and recognition of their type in broad applications.