Due to impact of radar measurement error and parameter space partition size, track initiation method based on parallel Hough transform still reveals a lot of clutter. To solve this problem, a parallel Hough transform track initiation algorithm based on fuzzy proposed. After mapping the membership degree of fuzzy logic function in different time into the parameter space separately, the accumulated results with the same index constitute an accumulated vector. According to the rules, the accumulated matrix can be obtained. Finally, we a decision on whether to initiate the track or not through the threshold detection. Simulation results show that this algorithm can constrain of mapping error effectively, decrease the occupancy of false tracks and initiate the track accurately under dense clutter environment.