11:30 〜 11:45
[PEM10-04] Extraction of ionospheric echoes for time series Ionograms
キーワード：イオノグラム、電離圏観測、動体検出
Sudden changes in the ionospheric environment can cause some radio disturbances and adversely affects our daily lives. Therefore, it is important for space weather forecast to make continuous observations of ionospheric environment in quasi-real time. Ionograms, which are a proxy for the height distribution of ionospheric electron density, are often used as a method of ionospheric observation. In addition to the ionospheric echo signal, ionograms often contain noises from measurements and computational processing. The purpose of this study is to develop a technique to generate ionograms that are robust to these noises and easy to read parameters of the ionospheric echoes.
In this study, we prepared 960 ionogram images (January 1, 2019, Sasaguri in Japan, FM-CW Radar, measured every 1 minute 30 seconds) and arranged them in time series order. We performed motion detection on the video images. Here, the motion detection method is based on the background subtraction algorithm with an additional penalty based on the echo characteristics of the appearance of high intensity noise in the frequency direction of the ionogram. As preprocessing steps, we set a threshold value for each ionogram based on the strength of the ionospheric echoes and removed weakly reflected echoes. Then, we perform motion detection on the denoised ionogram time series data to separate ionospheric echoes from background noise. In this result, we revealed that motion detection enables ionospheric echo extraction from noisy ionograms.
In this study, we prepared 960 ionogram images (January 1, 2019, Sasaguri in Japan, FM-CW Radar, measured every 1 minute 30 seconds) and arranged them in time series order. We performed motion detection on the video images. Here, the motion detection method is based on the background subtraction algorithm with an additional penalty based on the echo characteristics of the appearance of high intensity noise in the frequency direction of the ionogram. As preprocessing steps, we set a threshold value for each ionogram based on the strength of the ionospheric echoes and removed weakly reflected echoes. Then, we perform motion detection on the denoised ionogram time series data to separate ionospheric echoes from background noise. In this result, we revealed that motion detection enables ionospheric echo extraction from noisy ionograms.