SAR image channel extraction method combining gray-level threshold-value segmentation and contour shape identification

Abstract

The invention discloses an SAR image channel extraction method which combines gray-level threshold-value segmentation and contour shape identification. The method includes step1: adopting gray-level threshold-value segmentation to carry out a first background segmentation on an SAR image; step2: according to channel contour shape characteristics, carrying out sectioned modeling on channel areas, wherein the channel areas are represented as combination of a plurality of minimum external-connection rectangular windows; step3: combining minimum external-connection rectangular windows in the same area so as to form a plurality of second rectangular windows; step4: according to the shape and connectivity of a channel contour, splicing second rectangular windows which satisfy a condition into a rough channel area; and step5: carrying out gray-level threshold-value segmentation again so as to obtain a channel extraction image. The SAR image channel extraction method which combines gray-level threshold-value segmentation and contour shape identification inhibits background noises, which are approximate to both a water-body gray level and a shape characteristic in the SAR image, through combination of an image gray-level characteristic and a shape contour characteristic of the channel and a multi-mode SAR image segmentation decision so that channel areas are extracted precisely and accurately.

Description

technical field [0001] The present invention relates to the segmentation and recognition technology of Synthetic Aperture Radar (SAR, Synthetic Aperture Rader) river image, in particular to a method for extracting a river based on a SAR image using combined grayscale threshold segmentation and contour shape recognition technology, which belongs to computer vision technology field. Background technique [0002] Compared with optical imaging, synthetic aperture radar (SAR) imaging has unique advantages in river channel target recognition and extraction in remote sensing images, mainly reflected in: 1. The spatial resolution of synthetic aperture radar (SAR) is relatively high, which can describe river channels more accurately Target. In addition, it is capable of all-weather and all-weather observation of the earth and has a certain penetrating ability; second, the water body in the river channel area has significant electromagnetic wave reflection characteristics, which is r...

AI Technical Summary

Problems solved by technology

[0004] At present, the resolution of SAR images has been developed to the sub-meter level, and more and more small targets are likely to form strong background noise because their electromagnetic wave reflectivity is close to that of water bodies, and it is difficult to filter them out through the gray threshold

In addition, the gray-level threshold segmentation cannot distinguish the difference between the river channel area and the sporadic water body. A single grayscale threshold segmentation cannot effectively achieve suppression, thus forming serious interference in the segmentation results

Finally, the coherent speckle noise inherent in SAR images will also have a serious impact on channel extraction

[0005] The outline of the river channel is different from group targets such as tanks and aircraft, and also different from regular linear targets such as bridges and roads. The shape area is not straight, and the thickness of the river channel in different sections is also significantly different. In addition to the influence of artificial structures such as bridges and docks, the prior knowledge that the channel profile contains parallel line pairs is often not applicable. Therefore, for the channel profile shape Accurate modeling is very difficult, which will seriously affect the identification of river channels based on contour morphological features

At present, the Snake modeling method is commonly used to model the river channel contour, but because it is sensitive to the initial contour position and noise, its application is subject to certain restrictions

[0006] The river targets in SAR images are not suitable for extraction by advanced pattern recognition methods such as clustering, because the appearance of different rivers varies greatly and is relatively complex

According to the existing feature extraction and clustering methods, it is difficult to accurately cluster the river and non-river areas, let alone find the optimal classification surface

Images