Research of Semantic Understanding on Target Region of Interest for Fuzzy Image
QingE Wu,Weidong Yang,Zhiwu Chen,Ping Zhang
DOI: https://doi.org/10.1016/j.engappai.2014.09.005
IF: 8
2014-01-01
Engineering Applications of Artificial Intelligence
Abstract:Image understanding can provide an accurate and rapid target identification method for some military affairs, public security, finance and other departments. How to carry out the semantic understanding of scene of fuzzy image is a problem to be solved urgently in many departments. Moreover, research on this problem is still in an initial stage at home and abroad. For fuzziness, incompleteness and scene semantic understanding on fuzzy image processing, this paper studies a variety of fuzzy signals, analyzes the uncertainties classification and their influence, constructs a three-tier processing mode framework that can eliminate fuzziness processing, repair processing, dynamic combination processing, and presents some methods and algorithms for fuzzy signal processing. Through defining the support, this paper selects target region of interest (ROI), extracts some multidimensional and effective characteristics of targets on ROI, and builds a fuzzy recognition algorithm for targets. By defining the correlative importance between targets, as well as expert knowledge or associated experimental data, this paper carries out a semantic scene understanding on ROI, and presents a semantic understanding method for fuzzy image understanding. Using the combination method of simulation and instance experiments, this paper systematically analyzes the validity of the model and algorithms. The research developed in this paper can pave a new way to improve the real-time, speed and accuracy of information processing, have an important theoretical reference and practical significance, further form the theoretic base for uncertain information processing, and also provide a new idea and way on target recognition for a variety of scenarios.