Ridgelet bi-frame system-based SAR image airfield runway extraction method

Abstract

The invention discloses a ridgelet bi-frame system-based SAR image linear object detection method, which comprises the following steps of: (1) performing histogram equalization on an input SAR airfield image to be detected to obtain an enhanced result image; (2) pre-filtering the enhanced image and performing dyadic wavelet transform-based image edge detection on a result after the per-filtering; (3) setting a threshold value T1, setting the edges less than the threshold value in image edge detection results to be zero and reserving the edge regions more than the threshold value; (4) detecting the edge line of an airfield runway according to the reserved edge regions; (5) setting a threshold value T2 and performing binaryzation processing on detected airfield runway edge results; and (6) matching binaryzation processing results with the results of an original SAR airfield image after the filtering by using a linear tracking method, and determining the actual length of the runway. The method has the advantages of no complex post-processing process, high practicality, quickness, and precision positioning, and can be used for the extraction of the airfield runway in a relatively large complex SAR image scene.

Description

technical field [0001] The invention belongs to the technical field of image processing and relates to an image detection method for extracting airport runways in SAR images. Background technique [0002] Synthetic Aperture Radar, or SAR, has the ability to observe the earth's surface all-weather and all-weather, and with the improvement of signal processing technology, it has reached a very high spatial resolution, which makes SAR sensors in military reconnaissance and Battlefield awareness has played an increasingly important role. At present, SAR automatic target detection and recognition is a field that has attracted great attention at home and abroad. [0003] For SAR images, some important targets often contain a large number of linear features, such as ports, airports, buildings, etc. Therefore, the detection and extraction of line targets has always been an important issue in SAR image processing. [0004] Line targets in SAR images usually represent the boundaries...

AI Technical Summary

Problems solved by technology

This type of method has high requirements for the accuracy of low-level edge detection results. In the process of online primitive extraction, parallel line detection mostly uses hough transform, and hough has inherent defects in line detection: (1) it will be affected by raster image jaggedness Due to the influence of the problem, the result of hough transformation of a straight line is two or more straight lines; (2) only the longest local straight line can be detected; (3) only the position of the local straight line can be detected, and the straight line cannot be detected The endpoint coordinates of

Although this method can detect targets such as airport runways in relatively large and complex scenes, it still cannot solve the problem of airport extraction in large and complex SAR image scenes.

Images