Algorithms and integrated optoelectronic architectures for parallel image processing and adaptive optics applications are proposed and analyzed. Image processing is performed on the basis of an evolution equation with anisotropic gain that is introduced. The key components of the image-processing scheme are a coherent optical system for on-the-fly image-edge detection and an analog very large scale integration system for parallel implementation of the evolution equation with anisotropic gain. Examples of video data processing for edge enhancement, imaging through turbulent media, depth estimation from visual data, and motion tracking are presented.
© 1999 Optical Society of America
Original Manuscript: January 4, 1999
Manuscript Accepted: March 4, 1999
Published: July 1, 1999
Mikhail A. Vorontsov, "Parallel image processing based on an evolution equation with anisotropic gain: integrated optoelectronic architectures," J. Opt. Soc. Am. A 16, 1623-1637 (1999)