Abstract:

The dynamic programming (DP) algorithm for detecting dim targets is investigated. Due to the poor performance of detecting dim moving targets with traditional DP, an improved DP algorithm is proposed. The algorithm models the process of target state transition with a second-order Markov model, and on this basis transfers the optimization of traditional DP to a series of 2-dimensional optimizations. Thus, the target search region of the improved algorithm is narrowed on condition of considering a variety of target velocities, and the detection performance of dim moving targets for the DP algorithm is improved.