Multi-source data-based corn yield remote sensing estimation method

A corn yield remote sensing estimation method based on multi-source data relates to the field of agricultural remote sensing crop yield prediction, and comprises the following steps: obtaining and preprocessing long-time-sequence remote sensing data of a corn key growth period, and constructing a key growth period long-time-sequence mean value synthesis data set; obtaining multi-source remote sensing data influencing the corn yield, and extracting a regional corn planting range; constructing a machine learning model; further screening variable characteristics for modeling, and optimizing model parameters; evaluating the model precision; and drawing the regional corn yield spatial distribution. Based on a corn key growth period long time sequence image data set and actually measured sampling point data, in combination with Sentinel-2 satellite images, spectral indexes, environmental factors, crop parameters, soil attributes and other multi-source remote sensing data, a machine learning algorithm is utilized to