Remote sensing identification method of spring corn

Abstract

The invention discloses a remote sensing identification method of spring corn. The method is characterized in that the spatial distribution information of the spring corn in a regional range is extracted by using the differences between the remote sensing red light reflectivity, near infrared reflectivity and normalized difference vegetation index of the spring corn and those of other surface feature types and through classification system building, training sample selection, surface feature type remote sensing attribute index time sequence curve average extraction, identification feature selection, feature threshold determining and spring corn identification model building. The method has the advantages that discrimination rules can be built by selecting multiple remote sensing attribute indexes according to prior knowledge, then decision tree classification is used to identify the spring corn in the regional range, and high identification precision is achieved; the method is high in stability and universality and applicable to the operation implementation of large-range spring corn remote sensing extraction.

Description

technical field [0001] The invention relates to the technical field of agricultural remote sensing, in particular to a remote sensing extraction method for spatial distribution information of spring corn. Background technique [0002] Remote sensing has the characteristics of large-scale synchronous observation, high timeliness, and low cost. It can relatively conveniently obtain large-scale and full-coverage spatial distribution information of crops, and has been widely used to extract crop planting ranges. [0003] Crop remote sensing recognition methods include methods based on single-period remote sensing images and methods based on remote sensing time-series images. Different types of crops of the same season often have similar spectral characteristics in remote sensing images under large-scale conditions. Single-period remote sensing images generally have the phenomenon of different objects with the same spectrum, which cannot effectively distinguish crop types. The r...

AI Technical Summary

Problems solved by technology

[0004] The traditional method of identifying crops based on time series remote sensing images often only uses a certain remote sensing index, especially a certain vegetation index is the most common. This method has a weak ability to distinguish different crops with similar growth cycles and planting structures. , the tolerance to timing noise is low, which affects the accuracy of crop remote sensing extraction

Spring maize has a wide planting range and a wide ecological range, and there are many crop varieties with similar growth and development phenology characteristics. It is difficult to improve the remote sensing recognition accuracy of spring maize only by using a single remote sensing index

[0005] Although the normalized difference vegetation index NDVI is calculated from the red light and near-infrared bands, which reflects the advantages of information synthesis, and is often used as a single index for the identification of specific crop types, but this information synthesis may cover up a specific crop. The difference with other ground objects in the remote sensing red light or near-infrared band is not conducive to the identification of the specific crop type

Images