The Key Technology Study On The Surface Defects Detection Of Hot Heavy Rail

The heavy rail plays a decisive role in the national transportation.It's quality directly affects the safety of railway transportation. Hot heavy rail's surface quality inspection is the first step of heavy rail quality examination system.And it's also one of the most important steps, which has a positive effect on improving rail products quality, reducing scrap and enhancing economic efficiency. The speed, accuracy and other factors of this inspection is directly related to the overall quality of heavy rail.In this paper, with the situation of hot heavy rail inspection of a large iron and steel enterprise as research background, a suite of surface defects inspection on-line system was developed to overcome a variety of disadvantages of manual vision detection. Without high missed rate, low efficency, low accuracy.etc problems, this system can automatically examine the whole suface of the hot rail and output satisfied test results quickly.The page's main research contents are as follows:1. high-rate surface defects areas and focus detection area are analyzed together with the heavy rail geometry. Optimal position of the image acquisition devices is designed.2. Based on the analysis of lighting and radiation features and surface defects characteristics of hot heavy rail, suitable optical filter and lighting model is choosed.3. According to the features analysis of edge pixel variation of the heavy rail image, a strong contrast stretching algorithm is proposed to cut the background out from rail image.4. Aiming at the problem of unequal brightness of rail surface image, Matlab is used to simulate the lighting circumstance of hot heavy rail. According to the analysis results, a method is choosed to evaluate the unequal brightness function of rail image out and solve the problem.5. General characteristics of the hot heavy rail surface defects are analyzed and described. Further, two algorithms, image relevance examining between pixel lines algorithm and dark, overexposed areas examining algorithm, are proposed to test defects on the heavy rail surface. This combinded algorithm has highspeed, large adaptability for many kinds of surface defects, high-accuracy.etc advantages.