| With the rapidly development of Chinese economy, security of train control is broadly concerned as the train running faster and faster, and detecting flaw of train wheel and finding the deficient part accurately is always a very important problem to the national railway department and the transport department. Ultrasonic nondestructive detecting technology was widely used because of its penetrability, rapidity and other advantages. However, conventional ultrasonic imaging technology has difficulties in finding and evaluating the flaw accurately because of its low resolving capability.In order to improve the quality of ultrasonic imaging, enhance signal to noise ratio (SNR) and the ratio of flaw detection, a new image processing technique which combine B scan image of ultrasonic detection of wheels with algorithm for synthetic aperture focusing technique (SAFT) with the characteristics of high resolution is put forward basing on in-depth study of ultrasonic imaging and according to the methods and characteristics of ultrasonic detection of wheels.Firstly, made mathematical model of imaging arithmetic and data was calculated and analyzed amply, then the scheme of SAFT arithmetic was designed, the data was collected and image reconstruction was carried out in laboratory. The feasibility and effectiveness of the scheme was validated by analyzing results.Secondly, made the mathematical model of SAFT for wheel imaging of ultrasonic detection and analyzed data. The detection experiments are carried out on CRH type wheels which containing internal flaw. The image reconstruction of ultrasonic detection was achieved based on SAFT arithmetic, images are gained and the results are compared and analyzed.In addition, two optimized schemes were proposed in order to improve imaging quality, one is setting gate value and eliminating false flaw, the other one is the scheme based on correlation of ultrasonic signal. The schemes for arithmetic management were designed and the images were gained. Then compared the three arithmetic in imaging results, fore-and-aft signal and rapidity.Finally, using objective evaluation method for image quality, calculated the SNR of image reconstruction quantitatively. As expected, the SNR improves6～8dB, in addition, designed a software control interface which makes the image display more expediently and intuitively.The result show that the technique can effectively improve SNR and resolving capability of ultrasonic detection, which can provide more accurate information of flaw and enhance the rate of flaw detection. |
