Tbd Algorithm Based on Improved Randomized Hough Transform for Dim Target Detection

The track-before-detect (TBD) methodologies jointly process more consecutive scans and show superior detection performance for the low signal-to-noise ratio (SNR) targets over the conventional methods. A TBD algorithm based on improved Randomized Hough Transform for dim target detection is proposed in this paper. This algorithm uses the sequence numbers of scans to make sure that the point pairs are selected from different scans, avoiding the unreasonable situation that the point pairs may be selected from the same scan in the traditional Randomized Hough Transform (RHT). Second, it introduces a new voting method. Based on the minimum Euclidean distance criterion, this voting method finds the optimal parameter cell to vote, making the voting result better than the traditional RHT. In addition, we not only increase score of the optimal parameter cell but also update the corresponding parameter, thus suppressing the deviation between the recovered track and the target's track. Simulation results demonstrate the proposed algorithm can detect the dim target more rapidly and accurately than traditional RHT, especially under the background of low SNR.