Registration of motion-distorted interlaced images captured by a scanning vector imaging sensor

Area of Science:

• Computer Vision
• Image Processing
• Sensor Technology

Background:

• Images captured by cameras with scanning vector sensors, particularly those using interlaced schemes like staggered time delay and integration (TDI) cameras, are susceptible to motion and vibration distortions.
• These distortions degrade image quality, impacting subsequent analysis and applications.

Purpose of the Study:

• To develop and present an algorithm for realigning motion-distorted images acquired with interlaced scanning vector sensor cameras.
• To improve the quality of images degraded by motion during the exposure period, irrespective of the motion's nature.

Main Methods:

• The algorithm employs a two-step process: estimation of relative motion during exposure using normal least-squares estimation and subsequent image enhancement.
• It utilizes matrix computations, providing a computational advantage over traditional matching-based techniques.
• The method is designed to adjust image fields to correct for motion distortion.

Main Results:

• The algorithm successfully realigns image fields distorted by motion and vibrations.
• Demonstrated effectiveness on both simulated and real-world image data.
• The matrix computation approach offers improved efficiency compared to search-based methods.

Conclusions:

• The presented algorithm effectively corrects motion distortion in interlaced scanning vector sensor images.
• The method offers a computationally efficient solution for improving image quality degraded by motion.
• This technique has potential applications in various fields requiring high-quality image acquisition under dynamic conditions.