Search

Weather radar is an active remote sensing device used to monitor the full lifecycle changes in severe convective weather with high spatial and temporal resolution. Effective radar calibration is a crucial foundation for ensuring the high-quality application of observational data. This paper utilizes a UAV platform equipped with a high-precision RTK system and standard metal spheres to study the principles and methods of metal sphere calibration, constructing a complete calibration process and calibration accuracy evaluation metrics. Additionally, a collocated radar comparison observation experiment was conducted for cross-validation, and metal sphere calibration tests were performed on problematic radars. The experimental results indicate the following: (1) The combined application of a high-precision RTK system and a laser range camera can provide real-time position information on the metal sphere, improving the efficiency of radar target acquisition. (2) The calibration method based on UAV-suspended metal spheres can periodically conduct the quantitative calibration of Z and ZDR, achieving calibration accuracies within 0.5 dB and 0.2 dB, respectively, and supports the qualitative inspection of key parameters such as beamwidth and pulse width. (3) During field tests, a high success rate “coarse adjustment + fine adjustment + staring” sphere-finding technique was established, based on automatic switching between RHI, PPI, and FIX scanning modes. This method directs the UAV to adjust the metal sphere to the center of the radar distance bin, reducing the impact of uneven beam filling and bin crossing, ensuring the accuracy of scattering characteristic measurements.

With the progression of marine exploration and exploitation, as well as the advancements in mechanical intelligence, the utilization of the unmanned surface vehicle (USV) and the design of their guidance system have become prominent areas of focus. However, the stern ramp recovery of the USV is still in its infancy due to its unique attitude requirements and automation design. Furthermore, few studies have addressed the impact of maritime disturbances, with most research limited to simulations. To enhance the efficiency and accuracy of stern ramp recovery, this paper presents the development and construction of a novel recovery system. By incorporating physical modeling of disturbance forces acting on USVs at sea, the practicality of the system is improved. Additionally, an optimized genetic algorithm is introduced in the navigation module to improve convergence rates and subsequently enhance recovery efficiency. A line-of-sight (LOS) algorithm based on average velocity is proposed in this paper to ensure the attainment of unique attitude requirements and to improve the effectiveness of stern chute recovery. This paper provides a detailed description of the independently designed USV hardware system. Moreover, simulations and practical experiments conducted using this experimental platform are presented, offering a new solution for the USV’s stern ramp recovery.

Drowsiness poses a serious challenge to road safety and various in-cabin sensing technologies have been experimented with to monitor driver alertness. Cameras offer a convenient means for contactless sensing, but they may violate user privacy and require complex algorithms to accommodate user (e.g., sunglasses) and environmental (e.g., lighting conditions) constraints. This paper presents a lightweight convolution neural network that measures eye closure based on eye images captured by a wearable glass prototype, which features a hot mirror-based design that allows the camera to be installed on the glass temples. The experimental results showed that the wearable glass prototype, with the neural network in its core, was highly effective in detecting eye blinks. The blink rate derived from the glass output was highly consistent with an industrial gold standard EyeLink eye-tracker. As eye blink characteristics are sensitive measures of driver drowsiness, the glass prototype and the lightweight neural network presented in this paper would provide a computationally efficient yet viable solution for real-world applications.