Remote sensing image port ship detection method capable of multi-resolution hierarchical screening

Abstract

The invention discloses a remote sensing image port ship detection method capable of multi-resolution hierarchical screening. For a ship in a port in a large-size high-resolution optical remote sensing image, a coarse-to-fine method is adopted for processing the image, a sudden anomaly detection method is combined with a Harris corner detector (Harris Corner Detector) detection method, and the efficient ship-in-port detection method is established. The remote sensing image port ship detection method comprises the steps of: firstly, carrying out sudden anomaly and multi-angle rotational sudden anomaly detection on a sampled image of the large-size image, judging suspected ship areas obtained by adopting the two methods, and subjecting the judging results to regional expansion to obtain a suspected salient region of a port region; and secondly, extracting the suspected salient region in the image at the original resolution, carrying out sudden anomaly and multi-angle rotational sudden anomaly detection on the suspected salient region at the original resolution, detecting bow features of a suspected ship obtained by adopting the two detection methods by using the Harris corner detector method, returning ship coordinates obtained through detection to the original high-definition image for marking, and finally acquiring a ship mark in the original high-definition image.

Description

Technical field [0001] The invention relates to a remote sensing image processing method, in particular to a method for detecting ships in a port in the rapid and automatic processing of large-size and high-resolution optical satellite remote sensing images. Background technique [0002] In recent years, high-resolution optical remote sensing images have become a popular field of remote sensing image target detection. Among them, ship detection based on optical remote sensing images is the focus of research, especially the research on ship detection in port areas with relatively busy ship traffic has become a hot technology, and with the development of aviation technology and sensor technology, people Massive high-quality and high-resolution optical remote sensing images can be obtained. Therefore, slow manual recognition can no longer meet the needs of actual operations, and many automatic fast detection methods have been produced, such as: adaptive target detection threshold m...

AI Technical Summary

Problems solved by technology

Therefore, at this stage, a large number of ship detection methods based on optical remote sensing images are mostly concentrated on the more complex remote sensing images of the ocean background.

Even the detection methods involving ships in the port are mostly based on the detection of small-sized optical remote sensing images, and the detected candidate areas containing ships are used as the detection results, making the detection results insignificant.

Images