An apparatus and a method for detecting, classifying and tracking road users on frames of video data are provided. An object detection unit identifies an object in a frame of the video data. A pre-filtering unit determines whether the identifier associated to an object corresponds to another identifier associated to the same object in the same frame of the video data when the identifiers overlap. A tracking unit matches the identified object to a corresponding object in another frame of the video data, such that a path of the sequence of matched objects is tracked throughout the frames of the video data. A scenario unit determines a safety indicator that is a function of the risk indicators indicating interactions between road users.