Inteligent transportation system (ITS) is becoming a world wide solution for traffic problem. various source of information should support the ITS decision making, to name a few: social media, mobile agent, and Closed Circuit television or CCTV. In this paper we present a method to estimate vehicles speed using video processing in real time. Principal Component Analysis (PCA) is used to clasify vehicles. Kalman filter is harnessed to track and identify passing vehicles in real time. Then vehicle speed can be estimated via Euclidean Distance method. Speed accuracy obtained from ten video data, is in the ranges of 63 to 99.5%. The video data from this research is made available for public use.