Detecting and Mining Similarities, Differences and Target Patterns in Sequences of Images Using the PFF, LGG and SPNG Approaches
Abstract
In phase I the identification and significance of the problem was the mining images, especially sequences of images or video for detecting-extracting, fusing and recognizing differences, changes and associating patterns. These types of problems are difficult challenges in the image analysis and computer vision research community. These difficulties mainly due to the textural nature of the images and the possible noisy conditions during their capture. The recognition component was not a part of the phase I, but it belongs to phase II. We did it, however, in phase I in order to focus in phase II on the integration and real-time issues. Thus, for the achievement of the phase I tasks (objectives), we have developed and/or used several methods such as Pixel Flow Functions (PFF) (or projections), Segmentation, Local Global Graphs (L-G), Genetic Algorithms (GAs). Registration (or Mapping), Curve Fitting. Wavelets, Region Synthesis, Stochastic Petri-Nets (SPNs), and others. The efficient uses of these methods in a certain sequence has produced the desirable results for each of the tasks. Here we present each task and the sequence of methods involved for obtaining
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2004
- Accession Number
- ADA424553
Entities
People
- Despina Bourbakis