A KNOWLEDGE-BASED SYSTEM FOR SEA STATE RECOGNITION AND TARGET DETECTION USING A SINGLE MARINE RADAR SENSOR

摘要

The knowledge of the sea and his environment conditions is an important aspect for the safety of ships during development operations. Visual methods for target detections suffer serious limitations under critical weather conditions; in these cases it is desirable to have a robust sea state detector to provide current environment information. The echoes provided by the standard marine X-band radar can be used, besides target detection, to obtain information about current ocean conditions. In this work we present an approach that combines the typical radar signal processing with knowledge based signal processing to increase the performance of target detection in changing environments. The proposed approach carries out the recognition of sea clutter environment by means of the texture and energy analysis with the purpose of select the best detection algorithm for such conditions. The approach is based on the knowledge provided by the sensor and the heuristic knowledge of the expert. The sea texture recognition method, knowledge base and inference engine are described. The analysis of the results using simulated and real radar data are presented and discussed.