Early micro-damage and environmental noise makes it difficult to extract and identify damage information from monitoring data. In this study， a quantitative enhancement method of structural damage information is proposed based on Mahalanobis distance and empirical mode decomposition （EMD）. Firstly， structural health monitoring data are decomposed by EMD to obtain multi-order intrinsic mode function （IMF）， and the damage sensitive components are extracted by the energy variation of IMF. Secondly， the piecewise Mahalanobis distance cumulant （MDC） is used as the evaluation index of damage information quantitative enhancement， and the accumulative stop criterion is established according to the area of its probability density function. Finally， the applicability and effectiveness of the proposed method in structural damage information extraction and quantitative improvement are verified by numerical simulation and model test data. The results show that the energy transfer of each order IMF is obvious after the occurrence of structural damage. Selecting the IMF with more transfer energy than its own energy and the multi-order IMF with positive relative energy change rate as the damage sensitive component can achieve better damage identification than the original signal. The area change of MDC value probability density function can be used as the cumulative stop criterion to realize the effective identification of minimal damage.