a g-based 0 Multi-scale SAR image water segmentation method based on joint distribution with gamma

Abstract

The present invention adopts G 0 Joint distribution with Gamma to construct the global energy functional, and introduce the fast fuzzy C means (FFCM) algorithm to initialize the level set function when processing the SAR image at the top of the pyramid, and propose a multi-scale G-based 0 A new method for water body segmentation based on joint distribution with Gamma. The present invention is based on G 0 Variational level set segmentation based on joint distribution with Gamma, further improving the level set initialization method involved.

Description

technical field [0001] The invention relates to the field of SAR remote sensing, in particular to a G-based 0 Multi-scale high-resolution SAR image water segmentation method based on joint distribution with Gamma Background technique [0002] Synthetic aperture radar (SAR) has the advantages of all-day and all-weather, and can realize large-scale, real-time and dynamic ground object monitoring. Compared with optical sensors, it is not afraid of wind and rain, and has all-weather characteristics of penetrating clouds and fog, which is more suitable for monitoring and extracting flood disaster information. [0003] Due to the coherent speckle noise characteristics of SAR images, segmentation methods based on statistical models that introduce SAR statistical properties into segmentation methods are receiving more and more attention. Although the traditional segmentation method based on region splitting / merging can be combined with the SAR statistical model, it can only be bas...

AI Technical Summary

Problems solved by technology

[0007] The existing SAR image segmentation methods based on statistical models mostly use Gamma modeling, which cannot meet the needs of high-resolution SAR data processing, and the segmentation accuracy is low

Secondly, the existing methods mostly use binary segmentation results such as otsu to initialize the level set function. Using this method to initialize the level set function is very easy to cause the final segmentation result to be disturbed by the shadow area whose gray value is close to the water body, resulting in low segmentation accuracy.

Images