The invention provides an object micro Doppler feature separation and extraction method based on edge detection. The method comprises that smooth pseudo Winger-Ville processing is carried out on echo signal to obtain a time-domain distribution matrix of the echo signals; a time-domain distribution map of the time-domain distribution matrix is seen as a grayscale image, the contrast of the grayscale image is improved, the edge of the grayscale image is sharpened, and edge detection is carried out on the processed image; separation and boundary extraction are carried out on object micro Doppler features, corresponding to two components, in an edge detection result on the basis an edge gradient inertial principle; and the gravity center of energy distribution within the boundary is calculated in the time-domain distribution matrix, transient frequency change rule curves of the two components are obtained, and the object micro Doppler features are resolved. The object micro Doppler features can be separated when the echo signals are two-component signals, and the method is simple in calculation and high in practicality, has low requirements for object prior information, and overcomes disadvantages of a present component extraction method.