ASA 130th Meeting - St. Louis, MO - 1995 Nov 27 .. Dec 01

4pSP7. An adaptive threshold target detection and estimation algorithm.

Girish Chandran

Jules S. Jaffe

Marine Physical Lab., Scripps Inst. of Oceanogr., Univ. of Calif. at San Diego, La Jolla, CA 92093-0238

An algorithm to isolate multiple targets reflecting multiple correlated signals arriving at a receiver is described. The algorithm stems from a specific application, an active high-frequency, three-dimensional imaging system that is used to image zooplankton [Jaffe et al., ``FTV: A Sonar for Tracking Macrozooplankton in 3-dimensions'' (to appear in Deep Sea Research)]. The signals that are transmitted are designed to have flat cross-correlation properties. The receiver algorithm exploits this property of the transmitted signal set and works by iteratively isolating progressively weaker targets at each iteration step. An important feature of the structure is the ability to adapt the threshold of the detectors of different signals to targets of various strengths. The algorithm leads to a receiver structure which is a modification of the classical parallel matched filter bank to detect targets and estimate their parameters. Experiments were performed in a tank to test the receiver structure. Various objects were suspended in known configurations and then two signals were transmitted at them. The algorithm was successfully able to isolate the targets and estimate their positions and strengths. [Work supported by NSF Grant No. OCE 9421876.]