Vehicular communication technology (V2V, V2I, V2X, etc.) allows participating vehicles to actively exchange information about their surrounding traffic. By sharing the local field of view (FoV) perception along with their own instantaneous states, communicating vehicles can cooperatively broaden and improve their perceptions of the overall traffic which allows more effective maneuver planning and vehicle energy management. This paper outlines a framework for decentralized multi-vehicle cooperative perception where each vehicle broadcasts its perception information and also acts as an individual fusing node for the received data. The Bhattacharyya distance filter (BDF) is applied for the data association process, identifying and grouping a set of estimates potentially representing the same vehicle. The associated estimates are then passed down to a covariance intersection-based fusion scheme that yields a single fused state estimate and covariance matrix for each vehicle. A metric is adopted to assess the performance of the cooperative perception scheme. The workings of the proposed framework and its potential advantages over other solutions is illustrated via simulations of connected vehicles in highway traffic scenarios with different vehicle densities.