An Automatic Change Detection Method for Rice Rice Integrating Pixels and Objects

Abstract

The invention relates to an automatic rice change detection method integrating pixels and objects, comprising the following steps: establishing a MODIS water body index time series data set and a MODIS water body index time series curve; calculating the MODIS water body index time series curve of the start and end year by pixel Based on the similarity of the MODIS water index time series curve, the suspected change area is judged; the multi-period Landsat image of the starting year is integrated, and the image object is obtained through image segmentation; the suspected change area is superimposed on the image object to obtain the suspected change object in the study area ; Synthesize multiple Landsat images of the end year, and further subdivide the suspected change objects in the study area; for the suspected change objects in the study area after fine segmentation, use the Landsat remote sensing image features to carry out the rice mapping of the start and end years respectively, so as to obtain the research results. Distribution map of rice dynamic changes from the beginning to the end of the year. The invention has the characteristics of high degree of automation, simple and easy to use, good robustness, high classification precision and the like.

Description

technical field [0001] The invention relates to a rice automatic change detection method integrating pixel and object. Background technique [0002] In the field of remote sensing application research, the early research results mainly focused on the monitoring of land use and its changes. Since the 21st century, with the continuous improvement of the temporal and spatial resolution of remote sensing images, the depth and breadth of remote sensing applications have continued to expand. In recent years, research on crop monitoring using remote sensing images has attracted much attention. However, due to the spectral similarity between different crops and the spectral inconsistency caused by various factors such as phenological changes and variety differences of the same crop in different regions, remote sensing monitoring of crops is facing challenges. Conventional remote sensing monitoring methods are difficult to meet the actual needs of real-time and rapid automatic moni...

AI Technical Summary

Problems solved by technology

If the object-oriented classification method is applied to time-series remote sensing images, it will be difficult to deal with the problem that the category of some pixels inside the object established by the segmentation of a certain period of remote sensing images has changed.

Images