Interpretation Mode of Spaceborne SAR Based on Analysis of Cognitive Needs

The intelligent interpretation of spaceborne SAR images is an important technical field for future remote sensing applications. However, due to the imaging defects of SAR itself, it is very difficult to divide the image interpretation samples. Command and control, as an important application field of image interpretation, has many cognitive scenarios. Based on the analysis of image usage requirements, this article calculates the information value of corresponding target samples under different application purposes through text analysis of relevant SAR image scientific literature, technical reports, etc., and uses this as the selection threshold for image interpretation to improve the efficiency of image interpretation. At the same time, this paper also designed a SAR image interpretation mode based on graph neural network, replacing the attention mechanism module with knowledge calibration, and providing technical support for the in-depth application of satellite-borne SAR.