A Hybrid Method for NMR Data Compression Based on Window Averaging (WA) and Principal Component Analysis (PCA)

Abstract

Prior to the advent of nuclear magnetic resonance (NMR) data inversion, a common approach for handling the large amount of raw echo data collected by NMR logging was data compression for improving the inversion speed. A fast compression method with a high compression ratio is required for processing NMR logging data. In this paper, we proposed a hybrid method to compress NMR data based on the window averaging (WA) and principal component analysis (PCA) methods. The proposed method was compared with the WA method and the PCA method in terms of the compression times of simulated one-, two-, and three-dimensional NMR data, the inversion times of compressed echo data, and the accuracy of NMR maps created with and without compression. We processed NMR log data and compared the inversion results with different compression methods. The results indicated that the proposed method with a high compression speed and a high compression ratio can be used for NMR data compression, and its accuracy depended on the precompressed echo number, and it is obvious that the method have practical applications for NMR data processing, especially for multi-dimensional NMR.