Abstarct: There is an increasing demand for improvements in condition monitoring of transport infrastructure all over the world. Bridges are key components of transportation infrastructure and require such monitoring. Estimation of dynamic characteristics such as natural frequencies, mode shapes and damping of structures seems necessary for health monitoring and applying vibration control methods if needed. Due to the importance of the subject, bridge mode shapes are extracted from passing vehicle response and also local mean decomposition method is applied on vehicle response to assess natural frequencies. For this purpose, vehicle equipped with sensor passes the bridge and the acceleration response is recorded. Through interaction between the bridge and vehicle, the moving vehicle can be considered as both exciter and receiver. Theoretically it is assumed that bridge and vehicle properties are given and as vehicle passes through the bridge, interaction force is applied in contact point. Equations of bridge and vehicle motion with finite element method are used for extracting the interaction force. In this research the bridge is modeled as a simply supported beam using FE method and beam elements consist of four degrees of freedom (axial force is not considered), and the vehicle is modeled as a sprung mass. For extracting mode shapes, short time frequency domain decomposition method is applied on vehicle response. A comparison between results of the proposed method and real mode shapes extracted from finite element method is done using Modal Assurance Criterion. The amount of Modal Assurance Criterion using proposed method is one, showing accuracy of the method to find mode shapes. In second section, local mean decomposition method is used. This method is a time frequency analysis approach, which can decompose the signals into a set of product function (PF) components. In this section variational mode decomposition method is also investigated. Results of the proposed methods are compared with real frequencies; local mean decomposition method frequencies are within a margin of 2.29 percent of error and variational mode decomposition method frequencies are within a margin of 0.6 percent, showing the accuracy of proposed methods. Higher mode of frequencies are obtained using variational mode decomposition method so the method can be applied when higher mode of frequencies are of interest.