Small sample remote sensing image scene classification method based on double-flow structure

• Main Authors: MUNESHIGE, CHAI XINGHUA, LEI YAOLIN, HE WENZHI, LIU XIA, CHEN YANQIAO
• Format: Patent
• Language: Chinese; English
• Published: 03.10.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a small sample remote sensing image scene classification method based on a double-flow structure. The method comprises the following steps: firstly, dividing a remote sensing image scene classification data set into a training set, a verification set and a test set; then respectively constructing scene sets based on the training set, the verification set and the test set; then constructing a global flow network in a double-flow structure; then constructing a local flow network in a double-flow structure; then constructing a deep neighbor neural network model based on a double-flow structure based on the global flow network and the local flow network; training and verifying the deep neighbor neural network model based on the double-flow structure based on the scene set; and then testing the trained deep neighbor neural network model based on the double-flow structure based on the scene set. According to the method, a double-flow structure with two branches of the global flow network and