Since past decades, more people in India are demanding own vehicles. Population in India increases day by day and so transportation methods. Road accidents are increased especially on highways. Many peoples get injured and lost their life because they do not see the road sign boards and also not aware of the meaning of road sign boards. Proposed automatic road sign detection system enhances intelligent transportation system by providing information about road sign to the driver and keeps drivers view on road, which will minimize accidents on the road and saves the life of people. A proposed system identifies the road sign using image processing techniques. A system collects the video of road signs with the help of a camera which is mounted on moving vehicle. Image videos are converted into frames of the image at the rate of 25frames/sec. Image pre-processing techniques such as converting the RGB image to greyscale, resizing of the image are used. Images are enhanced using median filtering technique. To extract features of image speeded-up robust features (SURF) technology is used. The standard road sign images are stored in the database and for classification of features, linear support vector machine (SVM) is used. Results are tested on 10 videos of road signs and it shows that proposed method correctly detects normal, blur and partial view of road sign videos.