Real aperture radiometer calibration method based on deep learning network

• Main Authors: DOU HAOFENG, WU YUANCHAO, LIU SHUBO, SONG GUANGNAN, LI YINAN, LI HAO, DANG PENGJU
• Format: Patent
• Language: Chinese; English
• Published: 18.04.2023
• Subjects: CALCULATING; COLORIMETRY; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT,POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRA-RED,VISIBLE OR ULTRA-VIOLET LIGHT; MEASURING; MEASURING QUANTITY OF HEAT; MEASURING TEMPERATURE; PHYSICS; RADIATION PYROMETRY; TESTING; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR

The invention discloses a real aperture radiometer calibration method based on a deep learning network, and belongs to the technical field of microwave radiometer calibration. Two-point calibration is established on the basis of the linear relation between the input brightness temperature and the output voltage of the radiometer, and actually, the radiometer is not ideal linear. The physical temperature of core components such as an antenna and a noise source in a microwave radiometer can influence the output voltage of the radiometer, so that the calibration precision is reduced. In order to overcome the defects, the invention provides a real aperture radiometer calibration method based on a deep learning network. The calibration method disclosed by the invention comprises the following steps: obtaining the physical temperature of a core device of the radiometer; generating an original scene brightness temperature; generating an output voltage; constructing a data set; training a deep learning network; and v