The main contents of the project include image defogging and enhancement, smoke and flame signal detection and fireworks alarm system. Based on the improved CycleGAN、improved YOLOv4 and Beidou short message, the image defogging model, fireworks detection model and alarm location communication are realized respectively.
The main contents of the project include monitoring network layout, error correction and baseline calculation. Based on the ionospheric NLIM linear interpolation model and multi station fusion solution, the error correction and baseline solution are realized, and the tropospheric global prior model correction and orbit error precision ephemeris correction are realized.
Using Beidou Positioning to realize electronic fence, After the vehicle enters the gas station, the smart camera will automatically identify the vehicle information, and then judge whether the vehicle has been registered. If it is a member vehicle, the intelligent large screen next to the oil gun will directly display the vehicle information and refueling habits, including oil products, amount, etc., and automatically allocate the available oil gun. After refueling, the vehicle can leave directly, and the system will automatically deduct the payment according to the refueling amount. The owner can issue an electronic invoice at any time in the app or applet, which greatly facilitates the refueling service through AI.
Copyright © 2019 BUAA Mobile and Social Computing Research Group