This paper presents a reasonable gridding-parameters extraction method for setting the optimal interpolation nodes in the gridding of scattered observed data. The method can extract optimized gridding parameters based on the distribution of features in raw data. Modeling analysis proves that distortion caused by gridding can be greatly reduced when using such parameters. We also present some improved technical measures that use human- machine interaction and multi-thread parallel technology to solve inadequacies in traditional gridding software. On the basis of these methods, we have developed software that can be used to grid scattered data using a graphic interface. Finally, a comparison of different gridding parameters on field magnetic data from Ji Lin Province, North China demonstrates the superiority of the proposed method in eliminating the distortions and enhancing gridding efficiency.
