Research Of Ultrasonic Railway Inspection Data Compression Based On Wavelet Transform

An ultrasonic railway inspection data compression algorithm based on wavelet transform is studied in this paper. Ultrasonic testing technology is currently the most widely used non-destructive testing technology, has been applied in various fields of nondestructive testing, especially the rail flaw detection. The original railway NDT data is very large, and takes a lot of storage device. So research of ultrasonic railway inspection data compression is necessary.The characteristics and mathematical model of ultrasonic flaw detection signal are analyzed, the common data compression method is further discussed. Measured by the compression ratio and the relative root mean square error, the most suitable wavelet bases for data compression is studied, the compression scheme is developed. Then encoding is used to improve the quality of wavelet compression.The mathematical characteristics of the wavelet bases are compared. Wavelet bases through a preliminary screening according to the characteristics related to wavelet compression. Then remaining energy is introduced for a further screening of wavelet bases, 3 wavelet bases suitable for ultrasonic signal compression are obtained. Finally, through data compression experiments, Coif5 is the optimal wavelet base for ultrasonic data compression.Through data compression experiments, the compression ratio and relative error at each decomposition level are obtained. Since there is contradictory relation between compression ratio and distortion, the quantization threshold adjusted based on the variable-controlling method. After many compression experiments the optimal decomposition level is set to be 6. Then the wavelet compression ratio is 5.571%, and the relative root mean square error is 1.19%.Entropy encoding is introduced improve the quality of wavelet compression without any further distortion. Arithmetic coding, adaptive arithmetic coding, LZW coding and its improvement coding method is discussed. All this coding methods’coding speed and compression ratio is analyzed. Finally arithmetic coding is selected.Finally, integrated compression experiment on rail inspection data is performed. Better compression results are obtained with the combination method of wavelet compression and coding. The wavelet compression ratio is 2.919%, and the relative root mean square error is 1.19%, each parameters is better than the current ultrasonic compression method, where the wavelet compression ratio is 4.46%, and the relative root mean square error is 5.31%.