A vegetation classification method and device suitable for optical remote sensing satellite images

Abstract

The invention discloses a method and device suitable for classifying vegetation through optical remote sensing satellite images. The method includes the steps of firstly, carrying out region segmentation on the images to be tested to obtain segmentation vector regions, obtaining vegetation index values VIv of different pure vegetation pixels of the different segmentation vector regions and vegetation index values VIs of different pure exposed soil pixels of the different segmentation vector regions, and then obtaining the pointed fractional vegetation cover rather than calculating the fractional vegetation cover through fixed empirical values. Values of the different images are different, even in the same image, the VIs and the VIv of the regions with different soil types and different vegetation types are different, the fractional vegetation cover calculation error caused by the regionalism difference and the vegetation type difference can be reduced through the method and device, and the effect of improving the vegetation classification accuracy is achieved.

Description

technical field [0001] The invention relates to the field of optical remote sensing satellite images, in particular to a vegetation classification method and device suitable for optical remote sensing satellite images. Background technique [0002] Remote sensing images are image data that reflect the spatial distribution and spectral information of the earth's surface features and spectral information acquired by spaceborne or airborne sensors. It has the characteristics of wide coverage and short imaging period. With the development of remote sensing image classification technology, remote sensing image Classification techniques are increasingly used in vegetation monitoring. In remote sensing monitoring of vegetation, a particularly important aspect is the calculation of vegetation coverage (FVC). At present, the linear spectral separation (LSU) method is mostly used in the field of remote sensing to estimate the vegetation coverage of multispectral and hyperspectral ima...

AI Technical Summary

Problems solved by technology

Due to VI s with VI v It changes with time and space, and the values ​​of different images should be different, even for the same image due to the difference of soil type and vegetation type, VI s with VI v will be different, therefore, it is unscientific to only use fixed empirical values ​​to calculate vegetation coverage

Images