After analyzing the invalid mapping and the poor real-time performance in batching voting of the classical Hough Transform (HT) in parameter space mapping method, an improved HT detection algorithm was proposed on the basis of converging mapping of the aerial target velocity constraint sampling and online voting.The simulation results show that the improved target track detection algorithm achieves better performance while SNR is greater than 0.8 dB, and the target acquisition time varies with different SNR.The target can be captured quickly when SNR is high, and we can use time to exchange for the probability of detection when it is low.