Abstarct: Various methods are available to estimate the speed such as ultrasonic methods, magnetic methods, radar methods, infrared methods, laser methods, video sequence methods, GPS , …. In recent years novel approaches to determine the speed baxsed on the image to a static camera has been proposed baxsed on a single image taken by a stationary camera. In this approach speed of object is estimated, using parameters that include length blur and direction blur. Due to the relative motion between the camera and moving object for an extended period of camera exposure time, motion blur will occur in a region of the image corresponding to the dynamic object. In the scene for any fixed time interval, the displacement of the vehicle in the image proportional to the amount of blur caused by the imaging process. The displacement shows itself as blurring the moving object in the period that shutter camera is open. Shutter time is camera intrinsic feature. To determine the speed, exact determine the parameters of blur is important, so that whatever these parameters are estimated more accurately, the speed will be achieved more accurately. In this thesis with aid Radon transform and Fourier transform and combining and integrating these two methods, the new method will be presented that will be estimate more accurately the parameters of blur. This method has several advantages and estimates the length blur and the direction blur with high accuracy. For investigate this method, several synthetic image and real image has been used. Also high accuracy of this method relation than other methods and previous approach will be compared. This method is a recursive methods that calculating several numbers of default value, most accurate and most appropriate blur length and direction finally will be achieved. Accuracy of this algorithm for synthetic image and real image will be assessed.