Abstract: To accurately extract fault feature information under strong background noise， a multi-scale improved differential filter （MIDIF） is proposed for rotating machinery fault diagnosis. The rotating machinery vibration signal is decomposed into a series of multi-scale improved differential filter signals using MIDIF. In view of that the MIDIF filtered signals exhibit varying extents of validity in revealing fault features， a weighted reconstruction method using correlation analysis is proposed in which the weighted coefficients are counted and distributed to the corresponding MIDIF filtered signals to highlight the effective MIDIF filtered signals and weaken the invalid ones. The weighted coefficients are multiplied with the MIDIF filtered signals under different scales to produce transient impulse components. The fault types of rotating machines are inferred from the fault defect frequencies in the envelope spectrum of the transient impulses. The results show that MIDIF is more accurate in extracting fault features than multi-scale average combination different morphological filter （ACDIF） and multi-scale morphology gradient product operation （MGPO）， and that it provides an effective method for rotating machinery fault diagnosis.