Remote sensing identification method for abandoned land

Abstract

The invention provides a remote sensing recognition method for an abandoned land, and the method achieves the extraction of the abandoned land from a remote sensing image through the steps of image preprocessing, feature extraction, model training, reasoning, logic operation and the like. By means of the abandoned land extraction method, the accuracy of the extracted abandoned land can be greatly improved, the efficiency is greatly superior to that of a manual mode, and the problem that manpower is consumed in abandoned land extraction is solved. In addition, the LUNT model is superior to the UNET model, the good integrality of the model classification result can be guaranteed, and meanwhile, the number of plaques is small. Meanwhile, the generalization of the TRNET model is better than that of the UNET model, and the abandoned land extraction accuracy can be improved under the condition that the characteristic quantity is not increased.

Description

technical field

[0001] The invention relates to the technical field of remote sensing images, in particular to a technology for identifying abandoned land in the field of agricultural remote sensing. Background technique

[0002] Remote Sensing Image (Remote Sensing Image) refers to the film or photo that records the electromagnetic wave size of various ground features, such as aerial photos and satellite photos. Each remote sensing image contains corresponding geographic coordinates. In other words, each pixel in a remote sensing image can obtain its unique geographic coordinate position.

[0003] So far, the research on cultivated land abandonment in my country mainly relies on the collection of statistical data and household survey data for comparative analysis. Due to the gradual change, instability, and scattered spatial distribution of cultivated land abandonment, it is difficult to confirm its information and grasp the dynamic changes of abandoned land. Due to the c...

Images