The degradation of image resolution reduces the detection performance in aerial imagery because it generates a large number of small objects, and accurately detecting these small objects remains a challenge. Existing methods mostly use a Super-Resolution (SR) model to first obtain the SR image of the Low-Resolution degraded image ( $I^{LR}$ ) and then use this image as the input of the object detection (OD) network to solve this problem. However, this architecture that involves executing a complex SR network before the detector is time-consuming and makes it hard to achieve real-time model inference. To address this challenge, we propose a simple and effective rotated small object detection method, named End-to-end Super-Resolution enhanced real-Time rotated object Detector (ESRTMDet). Firstly, we design a lightweight Embedded feature map Super-Resolution Module (ESRM) embedded in the detection model to enhance and amplify the backbone output features, making the detection heads detect small objects more easily. Furthermore, we train a parallel SR network branch (PSRB) simultaneously that uses the backbone feature to restore a High-Resolution image ( $I^{HR}$ ). Through our proposed feature alignment loss ( $\mathcal {L}_{AL}$ ) and feature affinity layer (FAL), our PSRB effectively guides the feature map enhancement of ESRM. Finally, through end-to-end joint optimization of the detector and PSRB, the detection performance on $I^{LR}$ is significantly improved. Extensive experiments over DOTA and UCAS-AOD demonstrate that our method can achieve state-of-the-art (SOTA) results. Additionally, we discard our PSRB and use $I^{LR}$ as the input during inference, reducing the inference time-consuming of our model. Therefore, our ESRTMDet-X not only achieves 77.11% mean of Average Precision (mAP) on the degraded DOTA dataset, but also achieves an amazing inference speed of 337 FPS, thus obtaining the best speed-accuracy trade-off. The code and models will be released at https://github.com/liufeinuaa/aisodet.git .

Solar energy harvesting devices are widely used in smart agriculture nowadays. However, when lighting conditions are weak, such as through the night or on cloudy days, efficiency decays a lot. Additionally, as time goes by, more and more dust and bird droppings accumulate on the panel, which decreases the performance significantly. This paper aims to overcome the disadvantages mentioned above, and a novel wind–solar hybrid energy harvesting approach is proposed with an oscillation-induced dust-cleaning function. A wind-induced vibration device is specially designed in order to generate electrical energy and/or clean the photovoltaic panel. While in good lighting conditions, the device could keep the panel in a stable state and optimize the photovoltaic power generation efficiency. Such a hybrid energy harvesting approach is called a “suppress vibration and fill vacancy” algorithm. The experimental platform of the proposed device is introduced, and both experimental and simulation results are attained, which prove that using this device, we could realize multiple purposes at the same time.

The classically oriented object detection method often suffers boundary discontinuity and square-like problems, hindering the model’s ability to predict orientation accurately. Therefore, in this letter, we introduce a novel angle representation scheme named coupled dual-frequency phase-shifting coder (CDFP) which draws inspiration from optical measurement technology to address the aforementioned issues. Specifically, in the angle encoding stage, we represent the ground-truth (GT) rotation angle as a combination of two five-step phase-shifting at different frequencies and use this representation to supervise the learning of the model’s angle branch. In the angle decoding stage, alongside utilizing the corresponding dual-frequency phase-shifting decoding and unwrapping method, we impose additional constraints on the decoding angle range for predicted square-like objects. Extensive experiments on three challenging aerial image datasets using different detectors prove the effectiveness of our approach. Specifically, our RetinaNet-CDFP achieves an average improvement of 2.16% AP50 and 6.83% AP75 on DOTA, and when combined with RTMDet, our RTMDet-R-m-CDFP achieves state-of-the-art (SOTA) detection performance on DIOR-R and DOTA, with 70.11% and 78.77% AP50, respectively. Our codes will be released at https://github.com/liufeinuaa/aisodet.git .

This paper proposes a new piezoelectric shunt circuit based on bidirectional piezoelectric effect. Through the operation of injecting energy when the piezoelectric voltage is turned up and absorbing energy when it is turned down, it maximizes the energy efficiency of the control system and fully utilizes the electromechanical conversion potential of piezoelectric element. The working principle of the proposed method is introduced, the vibration attenuation of the structure under the condition of asymmetrical piezoelectric voltage is given. Finally, a high energy-efficient vibration control system is confirmed by experimental results, the structural vibration attenuation of 16.4 dB at 2.65 Hz is obtained under the power consumption of 465 mW.