Crop identification method and system based on convolutional neural network

Abstract

The invention relates to a crop identification method and system based on a convolutional neural network, and belongs to the field of intelligent identification. The remote sensing image and the crop type reference data are classified by using a classifier, the obtained crop type classification label is used for training a neural network model, and the trained neural network model can be used for efficient large-scale crop remote sensing automatic classification. According to the invention, manual labeling is not needed, a classification result of a traditional method with reliable precision can be directly adopted as a sample label, crop classification is carried out on a large number of remote sensing images, the tedious work of labeling the images is reduced, the cost is reduced, and the recognition efficiency and accuracy are improved.

Description

technical field [0001] The invention relates to the technical field of intelligent identification, in particular to a method and system for identifying crops based on a convolutional neural network. Background technique [0002] Driven by big data, deep learning has gradually become a research hotspot in the fields of remote sensing image target detection and semantic segmentation, and has great development potential. However, the deep learning method also has certain limitations. For crop classification, when training the convolutional neural network, it is necessary to provide a large number of labeled samples to be able to extract spatial features. To produce labeled samples requires a lot of manpower and material resources, which to some extent limits the application of deep learning methods in crop classification. Contents of the invention [0003] The object of the present invention is to provide a method and system for identifying crops with reduced cost and improv...

AI Technical Summary

Problems solved by technology

However, the deep learning method also has certain limitations. For crop classification, when training the convolutional neural network, it is necessary to provide a large number of labeled samples to be able to extract spatial features. To produce labeled samples requires a lot of manpower and material resources, which limits the application of deep learning methods in crop classification to a certain extent.

Images