SAR image-oriented aircraft target automatic detection method and system

Abstract

The invention discloses an SAR image-oriented airplane target automatic detection method and system, and the method comprises the steps: carrying out the airport detection of an airport SAR image, andobtaining an airport mask and an airport rectangular contour; performing aircraft detection on the airport rectangular contour of the airport SAR image to obtain a maskless aircraft detection result;and for the maskless aircraft detection result, removing false alarms by using an airport mask to obtain a final aircraft detection result. According to the method and system, the airport area can bequickly and accurately positioned, the false alarm rate of airplane detection is remarkably reduced, the calculation amount of subsequent airplane detection is reduced, and the airplane detection precision and speed are good.

Description

technical field [0001] The invention relates to an image target recognition technology oriented to SAR images, in particular to an automatic detection method and system for aircraft targets oriented to SAR images. Background technique [0002] Synthetic Aperture Radar (SAR) is an advanced active microwave earth observation equipment, and its imaging geometry belongs to the oblique range projection type. It has a certain penetrating effect and can effectively detect camouflaged targets. At present, the resolution of the high-resolution SAR system can reach the centimeter level, which can clearly image large-scale tactical targets and effectively help military personnel to identify targets. In addition, SAR also has the characteristics of all-weather and all-weather work, which is suitable for completing reconnaissance tasks under various harsh battlefield conditions. Since 1978, SAR has attracted countless attentions from the radar science community with its unique imaging ...

AI Technical Summary

Problems solved by technology

In 2015, R.Joseph and others proposed the YOLO algorithm, which is the first one-stage deep learning target detection algorithm. Compared with the two-stage detection algorithm, the detection speed is greatly improved, but the detection accuracy is reduced.

[0007] However, in view of the small size, different shapes, and complex backgrounds of aircraft targets in SAR images, the existing technologies generally have the problems of insufficient accuracy and high false alarm rate.

Images