Abstract: Rail inspection is an important task in railway maintenance. The speed and loads of trains have been increasing greatly in recent years, and these factors inevitably raise the risk of producing rail defects. Mainly the discrete surface defects impact the riding quality and safety of a railway system. However, it is a challenge to inspect such defects in a vision system because of illumination inequality and the variation of reflection property of rail surfaces. This project presents a real-time VIS for discrete surface defects of rail heads. VIS comprises the Image Acquisition Subsystem (IAS) and the image analysis subsystem. IAS acquires gray rail images for the surface of a rail head, and the latter processes rail images and detects possible defects. This paper propose the Local Normalization(LN) method to enhance the distinction between defects and background in a rail image, considering illumination inequality and the variation of reflection property of rail surfaces. VIS first acquires a rail image by the image acquisition system, and then, it cuts the sub image of rail. Track by the track extraction algorithm. Then, VIS enhances the contrast of the rail image .At last, VIS detects defects using the defect localization based on projection profile (DLBP), which identifies possible defects using the projection profile of the mean intensity over each longitudinal (or transversal) line. This is robust to noise and very fast. The proposed LN method and DLBP algorithm are better than the related well-established approaches. The distance of the crack can also be found based on certain assumption.