In this paper, an object detection and recognition method based on improved YOLOv5 is proposed for application on unmanned aerial vehicle (UAV) aerial images. Firstly, we improved the traditional Gabor function to obtain Gabor convolutional kernels with better edge enhancement properties. We used eight Gabor convolutional kernels to enhance the object edges from eight directions, and the enhanced image has obvious edge features, thus providing the best object area for subsequent deep feature extraction work. Secondly, we added a coordinate attention (CA) mechanism to the backbone of YOLOv5. The plug-and-play lightweight CA mechanism considers information of both the spatial location and channel of features and can accurately capture the long-range dependencies of positions. CA is like the eyes of YOLOv5, making it easier for the network to find the region of interest (ROI). Once again, we replaced the Path Aggregation Network (PANet) with a Bidirectional Feature Pyramid Network (BiFPN) at the neck of YOLOv5. BiFPN performs weighting operations on different input feature layers, which helps to balance the contribution of each layer. In addition, BiFPN adds horizontally connected feature branches across nodes on a bidirectional feature fusion structure to fuse more in-depth feature information. Finally, we trained the overall improved YOLOv5 model on our integrated dataset LSDUVD and compared it with other models on multiple datasets. The results show that our method has the best convergence effect and mAP value, which demonstrates that our method has unique advantages in processing detection tasks of UAV aerial images.

A Helmholtz resonator (HR) with an embedded aperture is an effective acoustic metamaterial for noise reduction in the low-frequency range. Its sound absorption property is significantly affected by the aperture shape. Sound absorption properties of HRs with the embedded aperture for various tangent sectional shapes were studied by a two-dimensional acoustic finite element simulation. The sequence of resonance frequency from low to high was olive, common trapeziform, reverse trapeziform, dumbbell and rectangle. Meanwhile, those HRs for various cross-sectional shapes were investigated by a three-dimensional acoustic finite element simulation. The sequence of resonance frequency from low to high were round, regular hexagon, square, regular triangle and regular pentagon. Moreover, the reason for these phenomena was analyzed by the distributions of sound pressure, acoustic velocity and temperature. Furthermore, on the basement of the optimum tangent and cross-sectional shape, the sound absorption property of parallel-connection Helmholtz resonators was optimized. The experimental sample with optimal parameters was fabricated, and its average sound absorption coefficient reached 0.7821 in 500–820 Hz with a limited thickness of 30 mm. The research achievements proved the significance of aperture shape, which provided guidance for the development of sound absorbers in the low-frequency range.

The variable noise spectrum for many actual application scenarios requires a sound absorber to adapt to this variation. An adjustable sound absorber of multiple parallel-connection Helmholtz resonators with tunable apertures (TA–MPCHRs) is prepared by the low-force stereolithography of photopolymer resin, which aims to improve the applicability of the proposed sound absorber for noise with various frequency ranges. The proposed TA–MPCHR metamaterial contains five metamaterial cells. Each metamaterial cell contains nine single Helmholtz resonators. It is treated as a basic structural unit for an array arrangement. The tunable aperture is realized by utilizing four segments of extendable cylindrical chambers with length l₀, which indicates that the length of the aperture l is in the range of [l₀, 4l₀], and that it is tunable. With a certain group of specific parameters for the proposed TA–MPCHR, the influence of the tunable aperture with a variable length is investigated by acoustic finite element simulation with a two-dimensional rotational symmetric model. For the given noise spectrum of certain actual equipment with four operating modes, the TA–MPCHR sample with a limited total thickness of 40 mm is optimized, which is made of photopolymer resin by the low-force stereolithography, and its actual average sound absorption coefficients for the frequency ranges of 500–800 Hz, 550–900 Hz, 600–1000 Hz and 700–1150 Hz reach 0.9203, 0.9202, 0.9436 and 0.9561, respectively. Relative to common non-adjustable metamaterials, the TA–MPCHR made of photopolymer resin can reduce occupied space and improve absorption efficiency, which is favorable in promoting its practical applications in the noise pollution prevention.

The limited occupied space and various noise spectrum requires an adjustable sound absorber with a smart structure and tunable sound absorption performance. The hexagonal acoustic metamaterial cell of the multiple parallel-connection resonators with tunable perforating rate was proposed in this research, which consisted of six triangular cavities and six trapezium cavities, and the perforation rate of each cavity was adjustable by moving the sliding block along the slideway. The optimal geometric parameters were obtained by the joint optimization of the acoustic finite element simulation and cuckoo search algorithm, and the average sound absorption coefficients in the target frequency ranges of 650–1150 Hz, 700–1200 Hz and 700–1000 Hz were up to 0.8565, 0.8615 and 0.8807, respectively. The experimental sample was fabricated by the fused filament fabrication method, and its sound absorption coefficients were further detected by impedance tube detector. The consistency between simulation data and experimental data proved the accuracy of the acoustic finite element simulation model and the effectiveness of the joint optimization method. The tunable sound absorption performance, outstanding low-frequency noise reduction property, extensible outline structure and efficient space utilization were favorable to promote its practical applications in noise reduction.

The objects in UAV aerial images have multiple scales, dense distribution, and occlusion, posing considerable challenges for object detection. In order to address this problem, this paper proposes a real-time multi-scale object detection method based on an improved YOLOv7 model (ATS-YOLOv7) for UAV aerial images. First, this paper introduces a feature pyramid network, AF-FPN, which is composed of an adaptive attention module (AAM) and a feature enhancement module (FEM). AF-FPN reduces the loss of deep feature information due to the reduction of feature channels in the convolution process through the AAM and FEM, strengthens the feature perception ability, and improves the detection speed and accuracy for multi-scale objects. Second, we add a prediction head based on a transformer encoder block on the basis of the three-head structure of YOLOv7, improving the ability of the model to capture global information and feature expression, thus achieving efficient detection of objects with tiny scales and dense occlusion. Moreover, as the location loss function of YOLOv7, CIoU (complete intersection over union), cannot facilitate the regression of the prediction box angle to the ground truth box—resulting in a slow convergence rate during model training—this paper proposes a loss function with angle regression, SIoU (soft intersection over union), in order to accelerate the convergence rate during model training. Finally, a series of comparative experiments are carried out on the DIOR dataset. The results indicate that ATS-YOLOv7 has the best detection accuracy (mAP of 87%) and meets the real-time requirements of image processing (detection speed of 94.2 FPS).

In order to achieve a balance between sound insulation and ventilation, a novel acoustic metamaterial of air-permeable multiple-parallel-connection folding chambers was proposed in this study that was based on Fano-like interference, and its sound-insulation performance was investigated through acoustic finite element simulation. Each layer of the multiple-parallel-connection folding chambers consisted of a square front panel with many apertures and a corresponding chamber with many cavities, which were able to extend both in the thickness direction and in the plane direction. Parametric analysis was conducted for the number of layers n_l and turns n_t, the thickness of each layer L₂, the inner side lengths of the helical chamber a₁, and the interval s among the various cavities. With the parameters of n_l = 10, n_t = 1, L₂ = 10 mm, a₁ = 28 mm, and s = 1 mm, there were 21 sound-transmission-loss peaks in the frequency range 200–1600 Hz, and the sound-transmission loss reached 26.05 dB, 26.85 dB, 27.03 dB, and 33.6 dB at the low frequencies 468 Hz, 525 Hz, 560 Hz, and 580 Hz, respectively. Meanwhile, the corresponding open area for air passage reached 55.18%, which yielded a capacity for both efficient ventilation and high selective-sound-insulation performance.