Stochastic gradient Bayesian SAR image segmentation method based on sketch structure

Abstract

The invention discloses a stochastic gradient Bayesian SAR image segmentation method based on a sketch structure, mainly used for solving the problem that SAR image segmentation in the prior art is inaccurate. The stochastic gradient Bayesian SAR image segmentation method comprises the following implementation steps of: (1), sketching an SAR image to obtain a sketch image of the SAR image; (2), according to an area chart of the SAR image, and dividing a pixel subspace of the SAR image; (3), performing hybrid aggregation structured surface feature pixel subspace segmentation through a method based on a stochastic gradient variational Bayesian network model; (4), performing independent target segmentation based on the sketch line aggregation feature; (5), performing line target segmentation based on a visual semantic rule; (6), performing segmentation of a pixel subspace in a homogeneous area by adopting a polynomial-based logistic regression prior model; and (7), combining segmentation results to obtain a segmentation result of the SAR image. By means of the stochastic gradient Bayesian SAR image segmentation method based on the sketch structure disclosed by the invention, the good segmentation effect of the SAR image is obtained; and the stochastic gradient Bayesian SAR image segmentation method can be used for semantic segmentation of the SAR image.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a synthetic aperture radar SAR (Synthetic Aperture Radar) image segmentation method based on a random gradient Bayesian of a sketch structure in the technical field of image segmentation. The invention can be applied to accurately segment different regions of the synthetic aperture radar SAR, and can be further used for target detection and recognition in SAR images. Background technique [0002] Synthetic aperture radar (SAR) is an important progress in the field of remote sensing technology, which is used to obtain high-resolution images of the earth's surface. Compared with other types of imaging technologies, SAR has a very important advantage. It is not affected by atmospheric conditions such as clouds, rainfall or heavy fog, and light intensity, and can obtain high-resolution remote sensing data all day and all weather. SAR technology has important guiding s...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that when obtaining the feature vector of the SAR image, the pixel-level features of the SAR image are used, and the unique structural features of the SAR image due to the correlation between pixels are not automatically learned. Insufficient use of structural features that truly represent the characteristics of SAR image features, resulting in inaccurate segmentation results

The disadvantage of this method is that when comparing and reasoning the structural features of the disconnected regions in the aggregation region, the reasoning network used in this method is the self-organizing feature map SOM network, because the self-organizing map SOM itself has artificial Determining the number of clusters, the shortcomings of long clustering time, and SOM will cluster the filter features with obvious direction differences into one class when processing the SAR filter features, which will greatly reduce the clustering accuracy and greatly affect The accuracy of SAR image segmentation

The disadvantage of this method is that when comparing and inferring the structural features of the disconnected regions in the aggregation region, this method uses the reasoning method of the self-organizing feature map SOM network, which needs to be manually determined. The number of clusters, and the clustering time is long, resulting in a decrease in clustering accuracy and affecting the accuracy of SAR image segmentation

The shortcomings of this method are that the boundary positioning of the aggregated area is not precise enough, the determination of the number of homogeneous areas is not reasonable enough, the regional consistency of the segmentation results is poor, and the independent target is not processed in the segmentation of the structural area.

Images