To track and detect dim targets more precisely and stably in less time

a dynamic programming trackbefore-detect（TBD） algorithm based on joint dual probability matrix is proposed.Firstly

by utilizing joint dual probability matrix algorithm

data of heterogeneous sensors are transformed into probability-values in probability matrices

which are used to be the input measurement-spaces of TBD algorithm.Then the target-track is obtained by a novel single-target dynamic programming TBD algorithm.Simulation results show that this algorithm is more suitable and stable for tracking and detection of dim targets by fusing multi-sensor measurements.