Optical Pattern Recognition for Missile Guidance.
Abstract
Progress on real-time spatial light modulators, image pattern recognition and optical signal processing for missile guidance is documented. A full description of our test and evaluation of the Soviet PRIZ spatial light modulator is inxcluded. In image pattern recognition, a unified formulation of four different and new types of synthetic discriminant functions is advanced. These include synthetic discriminant functions for intra and inter-class pattern recognition and multi-class pattern recognition. In the area of image pattern recognition, we also advance new statistical synthetic discriminant function filter concepts and a new principal component synthetic discriminant function. These analyses utilize new performance measures and new image models. Conventional holographic pattern recognition research conducted under AFOSR support is also reviewed. Our new AFOSR optical signal processing research concerns optical matrix-vecgtor processors. Initial research in this area includes fabrication of a fiber-optic microprocessor-based iterative optical processor and its use in adaptive phased array radar processing and for the calculation of eigenvalues and eigenvectors of a matrix. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 15, 1982
- Accession Number
- ADA130097
Entities
People
- David P. Casasent
Organizations
- Carnegie Mellon University