Abstract: Track before detect(TBD） procedures integrate a number of frames to improve the signal to noise ratio (SNR) of the radar echo, which could improve the performance of detection and track for weak targets. A novel generalized likelihood ratio testdynamic programming(GLRTDP) based TBD algorithm in rangeDoppler domain is proposed in the paper. Firstly, the target echo model is presented, and based on which, the expression of binary GLRT based TBD algorithm is derived, and after the data matrix is transformed to discrete rangeDoppler domain, DP algorithm is used to search for the maximum value of the expression which could decrease the operation quantity. Then, the performance of the proposed algorithm is analyzed by the probability of false alarm and the probability of detection. Finally, the algorithm is verified by simulation experiments, and the numerical results suggest that the performance of proposed algorithm is better than the traditional TBD detection method and it can be realized effectively to detect and track weak targetsin low signal to noise ratio background.