Software Design And Imaging Processing Algorithm Research Of Ultrasonic TOFD System

As an important way of nondestructive ultrasound detection,ultrasonic TOFD detection technology is widely used in industrial testing.Compared with other ultrasonic detection techniques,it has the advantages of high detection sensitivity,accurate defect location and the detection results are not affected by ultrasonic wave amplitude.The design and implementation of an ultrasonic TOFD detection software running on the Android platform is introduced.Because of the comprehensive reference of domestic and foreign products,the software has more rich functions and can provide more humanized services.The main content of the paper is as follows:1.Research on relevant principles of TOFD detection technology.It mainly includes the quantitative calculation of defect and imaging method.2.Software design of TOFD detection system.Based on domestic and foreign market product research and current research status,software requirements are planned and the corresponding technical route is developed.3.The implementation of key technologies.Some technologies are discussed,including Android platform communication connection based on USB2.0 and uses JNI interface to realize real-time and Offline imaging and Android layer overlay technology implemented with GLSurfaceView and SurfaceView.4.Research on TOFD image processing algorithm.The TOFD image processing overview of straight wave processing,image enhancement,pseudo-color mapping and image denoising is expounded.On the basis of K mean clustering,the improvement of rainbow code pseudo-color is put forward and a wavelet denoising algorithm which is based on improved threshold function is proposed.The experimental results show that the improved rainbow code mapping algorithm has good improvement effect and is more suitable for TOFD images.Finally,the complete realization process of software pseudo-color and image denoising function in Android platform is given.5.Software testing.Automation tools(Profiler,Monkey,Emmagee)were used to test the function,stability and performance of the designed software.The results showed that the basic functions of the software could be used normally and the pseudo-color mapping function of rainbow code based on k-means clustering had a good effect.The software memory utilization rate is stable below 5% and performs well in terms of performance.