Tracking Anisotropic Shapes In Digital Image Sequences At High Resolution

Tech ID: 29936 / UC Case 2018-848-0

Summary

Researchers led by Thomas Mason from the Chemistry and Physics Department at UCLA have developed a program that can identify, classify, and track the movement of different shapes in videos.

Background

Advancements in imaging and video technology have allowed us to image and record videos more quickly and with higher resolutions.  However, these improvements come at the cost of having very large image sets that can be hundreds of gigabytes large.  A video can be thought of as a large set of sequential 2D images.  Despite these improvements in imaging technology, the way we analyze and extract information from them has not. Analysis of videos and images is still done manually, which becomes nearly impossible for large data sets or for videos that need around the clock surveillance.  Manual data analysis suffers from fatigue, time intensity, and variance.  Accurate and automated analysis of objects in images and video will become an absolute necessity as these new imaging technologies become incorporated into everyday life.

Innovation

Researchers led by Thomas Mason from the Chemistry and Physics Department at UCLA have developed a program that can identify, classify, and track the movement of different shapes in videos.  Their program uses Mathematica to first identify objects of varying shapes.  It then uses key shape features to identify the same object from frame to frame to track its trajectory and rotation.  This type of software is especially useful for tracking cells or tiny organisms in a microscope setting.  The algorithm also does not depend upon computationally heavy functions, and thus gives it the advantage of being able to track trajectories and rotational movement in real time.  This makes it especially useful for applications like automated driving and for defense purposes.

Applications

  • Video and image analysis

- Cell tracking in light microscopy

- Cell/organism tracking in fluorescent micrsoscopy

  • Computer vision

- Automated driving

- Defense systems

Advantages

  • Can be done in real time
  • Flexible for different shapes
  • Accurate

Patent Status

Patent Pending

Contact

Learn About UC TechAlerts - Save Searches and receive new technology matches

Inventors

  • Mason, Thomas G.

Other Information

Keywords

segmentation, trajectory, rotation, tracking, anisotropic shapes, automatic detection, automated analysis, video analysis, image analysis, microscopy

Categorized As