Retina Project | Available Intellectual Property | Rensselaer Polytechnic Institute

Retina Project

RPI ID: 2002-042-201

Innovation Summary:
This technology presents an advanced algorithm for image registration using a dual bootstrap iterative closest point (ICP) method. It improves alignment accuracy between images by dynamically adjusting transformation models during iterations. The method is robust against noise and initial misalignments. It is particularly effective for medical imaging and computer vision applications.

Challenges / Opportunities:
Accurate image registration is critical in fields like medical diagnostics and remote sensing. Existing methods often struggle with convergence and precision. This invention offers a more reliable and adaptive approach to aligning complex image datasets. It provides an opportunity to enhance image-guided procedures and automated analysis systems.

Key Benefits / Advantages:
✔ High-precision image alignment
✔ Robust to noise and initial errors
✔ Adaptive transformation modeling
✔ Suitable for complex datasets

Applications:
• Medical imaging
• Computer vision
• Remote sensing
• Augmented reality

Keywords:
#Image registration #iterative closest point #computer vision #medical imaging #transformation modeling

Intellectual Property:
US Issued Patent 7,177,486
Patent Information:
Inventors:
Badrinath Roysam
Charles Stewart
Chia-Ling Tsai
Keywords:
LED
Lighting & Illumination Technology
For Information, Contact:
Natasha Sanford
Licensing Associate
Rensselaer Polytechnic Institute
sanfon@rpi.edu