The ionosphere is an important component of the near Earth space environment.The three common methods for detecting the ionosphere with high frequency(HF)radio signals are vertical detection,oblique detection,and obli...The ionosphere is an important component of the near Earth space environment.The three common methods for detecting the ionosphere with high frequency(HF)radio signals are vertical detection,oblique detection,and oblique backscatter detection.The ionograms obtained by these detection methods can effectively reflect a large amount of effective information in the ionosphere.The focus of this article is on the oblique backscatter ionogram obtained by oblique backscatter detection.By extracting the leading edge of the oblique backscatter ionogram,effective information in the ionosphere can be inverted.The key issue is how to accurately obtain the leading edge of the oblique backscatter ionogram.In recent years,the application of pattern recognition has become increasingly widespread,and the YOLO model is one of the best fast object detection algorithms in one-stage.Therefore,the core idea of this article is to use the newer YOLOX object detection algorithm in the YOLO family to perform pattern recognition on the F and E_(s) layers echoes in the oblique backscatter ionogram.After image processing,a single-layer oblique backscatter echoes are obtained.It can be found that the leading edge extraction of the oblique backscatter ionogram obtained after pattern recognition and image processing by the YOLOX model is more fitting to the actual oblique backscatter leading edge.展开更多
基金Supported by the National Natural Science Foundation of China(42104151,42074184,42188101,41727804)。
文摘The ionosphere is an important component of the near Earth space environment.The three common methods for detecting the ionosphere with high frequency(HF)radio signals are vertical detection,oblique detection,and oblique backscatter detection.The ionograms obtained by these detection methods can effectively reflect a large amount of effective information in the ionosphere.The focus of this article is on the oblique backscatter ionogram obtained by oblique backscatter detection.By extracting the leading edge of the oblique backscatter ionogram,effective information in the ionosphere can be inverted.The key issue is how to accurately obtain the leading edge of the oblique backscatter ionogram.In recent years,the application of pattern recognition has become increasingly widespread,and the YOLO model is one of the best fast object detection algorithms in one-stage.Therefore,the core idea of this article is to use the newer YOLOX object detection algorithm in the YOLO family to perform pattern recognition on the F and E_(s) layers echoes in the oblique backscatter ionogram.After image processing,a single-layer oblique backscatter echoes are obtained.It can be found that the leading edge extraction of the oblique backscatter ionogram obtained after pattern recognition and image processing by the YOLOX model is more fitting to the actual oblique backscatter leading edge.
