Concrete Structure's Ultrasonic Testing DSP And 2D CT Software Designing

The research in this paper is supported by the NSFC's(National Natural Science Foundation of China's) Funding: Smart Concrete and its Structure.( Program No:50238040.)At the beginning of this paper, the theory of scientific visualization tech and CT has been introduced. Several basic and important algorithms are explained and given comparison from the facet of matrix theory in Advanced Mathematics followed. The first-arrival traveling time of ultrasound transmitting various samples of concrete which contain different kinds of flaws are collected and used as the original projection data in 2D ultrasound CT imaging reconstruction. Later, the theory of neural network and wavelet analysis are adopted in imaging segment and ultrasonic signal's analysis respectively. The main tasks of this paper are listed below.1. The mathematical model of Ultrasonic CT testing are set up, iterative computing algorithms are adopted in imaging the concrete's inner condition. According to the traveling-time imaging theory, the mass sparse linear equations are built in the meshed computing process. Different reconstruction methods are adopted in solving such equations;the experiments results show that SVD method is more suitable to get a clearer ultrasonic CT imaging.2. The BP neural network method for segmenting the target and detecting the edge according to the pixel changing in different zone on the color image is also studied in this paper. The samples and their corresponding binary images for BP neural network training are derived from the ultrasonic CT images in concrete's non-destructive detecting. Because such network has good characters, its results show that the segmentation and the edge detecting effects of the BP network method are satisfying and also easy to design the network. Thus such intelligent segmenting method can be used in bulk information processing on ultrasonic imaging.3. Multi-scale wavelet analysis tech has been used to wipe off the noise of ultrasonic signal and distinguish the first-arrival time of ultrasound traveling. Through the decomposition of signal in different wavelet scale, the noise is largelyeliminated after recomposing the wavelet coefficient. Besides, Lipschitz index has been made used of to depict the signal's singularity to acquire the FAT by max modulus changing.The fruits achieved in this paper will make ultrasonic CT testing have better application in Concrete's NDT and have certain significance to guide's development. Authors deeply believe that ultrasonic computerized tomography technique will have vast prospect in application.