29-08-2023 дата публикации
Номер: CN116665070A
Принадлежит:
The invention discloses a terrace remote sensing recognition method fusing a jump network and a Mask R-CNN model, and the method comprises the steps: (1) obtaining a high-resolution remote sensing image for terrace recognition, and carrying out the preprocessing, including radiometric calibration, atmospheric correction, geometric correction, wave band fusion and image cutting; (2) artificial interpretation is carried out according to the preprocessed high-resolution remote sensing image, a terrace vector data sample library is obtained, and a data set for model training, verification and testing is made; (3) fusing the hopping network with a Mask R-CNN model, and proposing a JAM-R-CNN model; and (4) performing model training by using the training data set, selecting optimal parameters of the model by using the verification data set, and finally performing terrace identification and result evaluation by using the test data set. According to the method, the adhesion phenomenon of terrace ...
Подробнее