A HIERARCHICAL APPROACH TO DATA FUSION

THE GOAL OF THIS RESEARCH PROJECT IS TO DEVELOP A PRELIMINARY PROTOTYPE OF A HIERARCHICAL SYSTEM FOR DATA FUSION. THE SYSTEM WILL UTILIZE BOTH THE QUALITATIVE AND JUDGEMENTAL REASONING PROCESS OF A HUMAN EXPERT, AS WELL AS THE QUANTITATIVE CAPABILITIES OF MOST STATISTICAL AND PROBABILISTIC TECHNIQUES. ONE OF THE IMPORTANT AND DIFFICULT TASKS IS TO INTEGRATE CONFLICTING INFORMATION FROM VARIOUS SENSORS FOR SITUATION ASSESSMENT. OUR SYSTEM CONSISTS OF FIVE DISTINCT LAYERS. THE LOWEST LAYER, LEVEL 0, CONSISTS OF RAW DATA. LEVEL 1 IS FOR DATA ALIGNMENT AND DATA ASSOCIATION. LEVEL 2 IS FOR DATA CORRELATION AND FUSION. THE RESULTS FROM LEVEL 2 WILL BE USED BY LEVEL 3 FOR DATA ANALYSIS AND INTERPRETATION. IT IS COMMONLY CALLED SITUATION ASSESSMENT. THE RECOMMENDATIONS OF THE SYSTEM WILL BE PRESENTED TO THE USERS, WHICH IS THE HIGHEST LAYER - LEVEL 4.