Fast and precise object detection for hgigh-resolution aerial images has been a challenging task over the years. Due to the sharp variations in object scale, rotation, and aspect ratio, most existing methods are inefficient and imprecise. In this paper, we propose a different approach polar method. We locate an object by centrepoint, direct it by four polar angles, and measure it by polar ratio system. Our polar coordinate-based method, PolarDet, is a faster, simpler, and more accurate one-stage object detector. Also, our detector introduces a sub-pixel centre semantic structure to further improve classifying veracity. PolarDet achieves nearly all state-ofthe-art (SOTA) performance in aerial object detection tasks with faster inference speed. In detail, our approach obtains the SOTA results on authoritative remote sensing object detection datasets DOTA, UCAS-AOD, and HRSC2016 with 76.64% mAP (mean average precision), 97.01% mAP, and 90.46% mAP respectively. Most noticeably, our PolarDet gets the best performance and reaches the fastest speed (32fps) at the UCAS-AOD dataset.

No abstract

In the present study, the machine learning algorithm is utilized for the first time to improve the probe diagnosis. Machine learning methods are utilized to improve the Langmuir probe diagnostic accuracy and the diagnosable plasma parameter range without changing the probe structure based on the Langmuir probe. They provide a new way for experimentally obtaining electron density. A DC glow discharge simulation model and experimental equipment are established. Utilizing the discharge pressure and voltage as independent variables, the simulation and experimental electron densities are collected, the simulation and experimental data are utilized for training, and the plasma electron density outside of the pressure and voltage range of the training data is predicted, thereby achieving the prediction. Simultaneously, when the data amount is large enough, even without experimental measurement, the electron density can be obtained directly through the input parameters, without relying on the plasma physical model.

No abstract

No abstract