Recognition method of dry land crop based on remote sensing time series data

Abstract

The invention relates to a recognition method of a dry land crop based on remote sensing time series data. The method is based on the vegetation index of a research area and the time series data of a SWIR (short wave infrared band) to obtain the crop growing period in each growth cycle so as to deduce crop early growth stage and crop late growth stage to in turn establish crop early growth stage SWIR and EVI incremental product index and crop late growth stage SWIR and EVI incremental product index to comprehensively form whole crop growth stage SWIR and EVI incremental product index, and finally according to the whole crop growth stage SWIR and EVI incremental product index, the dry land crop can be identified. According to the method, by full use of the characteristics that in the whole crop growth stage of the dry land crop, SWIR band varies greatly, the SWIR and EVI are opposite in change direction, the SWIR band of a water crop varies slightly, and the SWIR and the EVI are relatively consistent in the change direction, the whole crop growth stage SWIR and EVI incremental product index is designed, and the method is used for identifying the dry land crop, and has the characteristics of good robustness, high classification accuracy, strong automation and strong anti-interference ability and the like.

Description

technical field [0001] The invention relates to the field of remote sensing information processing, in particular to a dryland crop identification method based on remote sensing time series data. Background technique [0002] Accurately and quickly grasping the planting area of ​​crops and their spatial distribution is crucial to ensuring food security. Bulk dryland crops such as wheat, corn, and potatoes are widely distributed and play an important role in my country's agricultural production. Other dryland crops such as peanuts, cotton, rapeseed, soybeans, sesame, etc., are closely related to people's lives as oil crops, feed or textile raw materials. Therefore, it is of extraordinary significance to quickly and automatically monitor the spatial distribution of dryland crops. The traditional agricultural sampling survey method is difficult to break through the limitations of high human and financial resources cost, time-consuming and incapable of full coverage. Remote s...

AI Technical Summary

Problems solved by technology

However, in the crop remote sensing monitoring business of my country's agricultural information remote sensing monitoring system, there are two issues worthy of attention: (1) In the identification of main crops, it is usually necessary to use cultivated land mask data, due to the lack of land use / cover data Problems such as timeliness and accuracy will inevitably bring errors, which directly affect the accuracy of remote sensing estimation of crop area; (2) Due to the frequent cloud and rain weather during the growing period of crops, remote sensing images in some periods are inevitably affected by clouds. The time-series curve of the remote sensing index is disturbed, which brings challenges to the remote sensing classification method based on multi-period or time series

Images