Multi-scale rotating ship target detection algorithm

Abstract

The invention provides a multi-scale rotating ship target detection algorithm. The algorithm comprises the following steps: acquiring a multi-scale feature map of an input image, wherein the multi-scale feature map comprises a first-class scale feature map, a second-class scale feature map and a third-class scale feature map of which the scales are sequentially increased; performing feature fusionon the multi-scale feature map by adopting a feature pyramid network, wherein the feature pyramid network adopts a ResNet residual network as a basic framework; inputting the feature map output by the feature pyramid network into a region suggestion network through a 3 * 3 convolution layer, classifying each anchor frame by adopting the region suggestion network according to a set classificationjudgment condition, endowing each anchor frame with parameter coordinates, and obtaining a rotating boundary frame based on parameter coordinate regression; performing adaptive region-of-interest alignment on a rotation bounding box generated by the region suggestion network to obtain a high-quality feature map; and screening the candidate boxes of the high-quality feature map according to a set rotation non-maximum suppression constraint condition to obtain a detection target.

Description

technical field [0001] The invention relates to the technical field of ship detection, in particular to a multi-scale rotating ship target detection algorithm. Background technique [0002] With the development of remote sensing technology, it is becoming easier to obtain high-resolution remote sensing images. For a long time, automatic detection of ships has played an important role in the field of remote sensing, and great progress has been made in port management, cargo transportation, and sea rescue. At the same time, the ship's berthing and navigation direction information is also of great significance. However, the feature of large aspect ratio makes ship detection more difficult than other object detection (such as vehicles, buildings, airplanes, etc.). [0003] In recent years, deep learning has achieved great success in the field of computer vision. The scheme based on regional convolutional neural network (Region-CNN, RCNN) provides a good way for target detecti...

AI Technical Summary

Problems solved by technology

[0006] In view of the fact that there are a large number of redundant areas in the ship target detection method based on the horizontal area, resulting in inaccurate detection results, and the problems of inaccurate coordinate regression and sample misclassification in the existing ship target detection method based on the rotation area, the present invention provides A Multi-Scale Rotating Ship Target Detection Algorithm

Images