Research progress on data preprocessing methods for geophysical multi-information fusion

Multimodal information fusion technology is an emerging field that has flourished in recent years, representing the application of artificial intelligence theories and technologies in information analysis and processing. It plays a crucial role in various domains, such as battlefield situational awareness, industrial robotics, remote medical care, and autonomous driving. With the rapid development of geophysical exploration theories and technologies, particularly with the significant advancements in China's airborne and satellite-based Earth observation technologies in recent years, the massive data generated poses a severe challenge to the traditional analysis models, which are primarily based on human experience. There is an increasing demand to introduce multi-information fusion technology into the field of geosciences, especially in the domain of geophysical exploration data. However, due to the differences in methods, scenarios, and equipment in geophysical observations, the data obtained have varying spatial distribution standards, making subsequent information fusion calculations difficult. Therefore, it is necessary to preprocess the observational data according to a unified standard to ensure that the data have consistent observational density. Currently, both the theory and technology for voxel-based preprocessing of observational data, which is crucial for multi-information fusion, are lacking. In some studies, traditional prediction techniques have been used to perform quasi-three-dimensional (or quasi-voxel-based) data standardization, but the effectiveness of these approaches is still under evaluation. Therefore, this paper reviews the existing data standardization preprocessing methods from the perspective of geophysical multi-information fusion methods. It introduces a semi-airborne electromagnetic observation case to discuss the performance of various existing methods, providing a methodological basis for further developing and refining the geophysical multi-information fusion processing technology system.