Interframe Image Processing with Application to Target Detection and Tracking.
Abstract
The restoration of images and the enhancement and detection of targets in cluttered background are the subjects of this research. The statistical approach is used in order to exploit temporal as well as spatial image redundancies. The images are modeled as a homogeneous random field. An autocorrelation function and a method of parameter identification are proposed. Experiments with several pictures are presented to validate the model. An analysis of two-dimensional recursive filters is presented. A three-dimensional recursive filter is developed which exploits the spatial as weell as the temporal image redundancies. A class of hybrid filters is proposed which improves the performance of the recursive filters. Several experiments with pictures are presented to show the ability of the hybrid filters in picture restoration.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1979
- Accession Number
- ADA068545
Entities
People
- Jose Leite Pereira Filho
Organizations
- Naval Postgraduate School