Remote sensing images accumulate over time, but there are still some missing situations. The existing remote sensing images can be used to reverse the missing remote sensing images, supplement the database and provide information for related research. The characteristics of remote sensing image data are analyzed, and the methods and ideas for deducting remote sensing images are envisaged. The existing algorithms and existing cases are used to study and think about related algorithms, which provides reference and ideas for further research. With the development of aerospace technology and the continuous breakthrough of sensor levels, the acquisition of remote sensing image data has become easier, and the image quality has gradually increased. Remote sensing image data has accumulated over the years. For a research area, there are often references to remote sensing image data from a variety of sensor sources, and the shooting time is usually more than one. These multi-temporal multi-source remote sensing image data have important reference value for the study of natural phenomena, environmental changes, human activities, and topography. However, due to the limitation of shooting, such as the period of the satellite, the influence of the weather, the uncertainty of the aviation platform, etc., there are still omissions in the long-term accumulation of remote sensing image data, and it is difficult to achieve continuous sensing of multiple sensors in a certain area without interruption. There will be a lack of remote sensing data at a certain time in the study area. When analyzing the research area, the remote sensing data corresponding to the determined time is the most important research data, and it can be used to obtain accurate information. The lack of direct remote sensing image data may lead to inaccurate research results or follow-up The processing and analysis are interrupted, and there is an urgent need for a method to make this data available for restoration and retrieval. However, time has passed, and accurate remote sensing data at that time will never be available. Based on the above problems, this paper proposes a concept of remote sensing image data acquisition, which can supplement the database and provide a reference for remote sensing data for scholars who need this information. And think and expand. The main idea is to simulate and deduct some type of remote sensing image data that has not been acquired at certain time points through existing remote sensing image data and other data. At the same time, this paper also proposes some new ideas to solve related problems. The Significance and Feasibility Analysis of Deducing Remote Sensing Image Data Depending on the sensor, remote sensing images can be divided into high-resolution full-color images, multi-spectral images, hyperspectral images, SAR images, near-infrared images, etc. The general remote sensing platform will be equipped with one or several sensors, but cannot guarantee Sensor diversity of remote sensing images in fixed time fixed regions. The imaging platform for remote sensing images is generally an aerospace and aviation platform. The aerospace platform has a high working height, so when the lower altitude weather changes (such as the emergence of clouds), the acquisition of images is limited; the re-visiting period of non-geostationary satellites makes it impossible to conduct research on the study area. Uninterrupted image acquisition on the phase; aviation platforms mainly include various reconnaissance planes and drones, and their remote sensing image acquisition is more affected by tasks and artificial control. In summary, the current remote