Detecting motion through dynamic refraction

Area of Science:

• Optics and Photonics
• Computer Vision
• Robotics

Background:

• Refraction creates dynamic distortions in atmospheric and underwater viewing conditions.
• Submerged animals and humans (e.g., divers, submarines) face challenges detecting objects through these distortions.
• Complete scene recovery from dynamic refractions is computationally difficult.

Purpose of the Study:

• To develop a method for detecting moving objects despite strong, random refractive distortions.
• To avoid the complex task of full scene recovery when only object motion is of interest.
• To enable detection of moving objects even with zero mean motion.

Main Methods:

• Utilizing a simple motion feature to discriminate moving object points from background.
• Leveraging the mutual independence between object motion and distortion.
• Focusing on detecting motion rather than inverting distortions.

Main Results:

• Moving objects can be detected with high accuracy and low false-positive rates.
• The method is effective even when distortions are severe and dominate object motion.
• Detection is successful even for objects with zero mean motion.

Conclusions:

• A straightforward motion-based feature allows for robust detection of moving objects through refractive distortions.
• This approach simplifies the problem by focusing solely on motion detection, bypassing full scene restoration.
• The findings have implications for underwater surveillance, robotics, and augmented reality systems.