A device and method for improving the identification accuracy of fetal heart rate deceleration. The method includes: collecting fetal heart rate data (101) performing baseline identification on the collected fetal heart rate data (102) preprocessing the collected fetal heart rate data (103) performing deceleration identification on the preprocessed fetal heart rate data according to preset deceleration judgment standards and a fetal heart rate data baseline to obtain a decelerated data segment (104) and calculating a deceleration attribute value of each decelerated data segment (105), and outputting the decelerated data segment and calculation result (106). The present invention can effectively eliminate the baseline variation part by employing methods such as dynamic threshold area method, continuous wave peak detection and hybrid deceleration detection and so on, and accurately identify each deceleration and the type thereof, avoiding the defects that the existing method easily causes large deceleration identification error but cannot identify continuous deceleration or eliminate baseline variation and cannot identify hybrid deceleration and so on. The solution in the present invention is not only suitable for the situation where deceleration occurs independently and is typical deceleration but also suitable for the situation where continuous deceleration and hybrid deceleration occur.
