Surface deformation classification method and device based on SAE and CNN models, and storage medium

• Main Authors: LI MENGHUA, YANG MENGSHI, WU HANFEI, TANG BOHUI
• Format: Patent
• Language: Chinese; English
• Published: 26.05.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention relates to a surface deformation classification method and device based on SAE and CNN models, and a storage medium, and the method comprises the steps: collecting SAR image data of a multi-scene time sequence in a target region, and obtaining an original time sequence image data set; performing data preprocessing on the original time sequence image data set by using a multi-temporal InSAR image data processing method to obtain a deformation displacement time sequence data set; classifying data in the deformation displacement time sequence data set to obtain a plurality of classified deformation displacement time sequence data sets; performing feature extraction on each classified deformation displacement time sequence data set by using an SAE network model to obtain a plurality of classified feature data sets; training a CNN model by using all the classification feature data sets to obtain an earth surface deformation classification model; performing deformation trend classification on the time