Forward collision warning (FCW) is a critical technology to improve road safety and reduce traffic accidents. However, the existing multi-sensor fusion methods for FCW suffer from a high false alarm rate and missed alarm rate in complex weather and road environments. For these issues, this paper proposes a decision-level fusion collision warning strategy. The vision algorithm and radar tracking algorithm are improved in order to reduce the false alarm rate and omission rate of forward collision warning. Firstly, this paper proposes an information entropy-based memory index for an adaptive Kalman filter for radar target tracking that can adaptively adjust the noise model in a variety of complex environments. Then, for visual detection, the YOLOv5s model is enhanced in conjunction with the SKBAM (Selective Kernel and Bottleneck Attention Mechanism) designed in this paper to improve the accuracy of vehicle target detection. Finally, a decision-level fusion warning fusion strategy for millimeter-wave radar and vision fusion is proposed. The strategy effectively fuses the detection results of radar and vision and employs a minimum safe distance model to determine the potential danger ahead. Experiments are conducted under various weather and road conditions, and the experimental results show that the proposed algorithm reduces the false alarm rate by 11.619% and the missed alarm rate by 15.672% compared with the traditional algorithm.

This paper researches the fabrication of a WR2.8 terahertz rectangular waveguide operating at the frequency ranging from 260 GHz to 400 GHz via UV-LIGA technology (UV-lithography, electroplating, and molding). In the process, megasonic agitation is applied to improve the mechanical properties and internal surface roughness of the WR2.8 rectangular waveguide. The effects of process parameters on the properties of structures are discussed, and optimized parameters are obtained to achieve accurate geometry dimensions. In addition, the highly crosslinked SU-8 is reliably removed from structures without damage through a synthesis method. The accuracy of the height and width of the WR2.8 rectangular waveguide is 5 µm and 2 µm, respectively, and the measured internal surface roughness is 79.6 nm. Moreover, experimental measurements and numerical simulations of the waveguide are conducted, and the difference between the two is also highlighted.

In this work, a Spindt cold cathode electron gun with a PPM (periodic permanent magnet) focusing system for a terahertz TWT (traveling wave tube) was designed and simulated based on the Pierce electron gun structure. More specifically, a new 3D (three dimensional) emission model was used, where the cathode radius of the electron gun was 1 mm and the cathode current was 30 mA, with an emitting half angle of about 28°. It was demonstrated that the electron beam was well focused with an electron beam radius of 0.3 mm and a filling ratio of 0.5 when the maximum value of the PPM field along with the axis was 0.122T. According to the simulation results, a planar cold cathode electron gun was developed. Measurements demonstrated that the I/V characteristics of the cold cathode gun were consistent with that of a cold cathode, revealing that the electrons emitted from the cathode are not intercepted when passing through the electron gun.

This article focuses on fabricating a 340-GHz folded waveguide (FWG) at frequencies ranging from 335 to 345 GHz using two-step UV-Lithographie Galvonoformung Abformung (LIGA) technology. The accuracy of the height and width of the 340-GHz FWG is ±5 and $\pm 1 \mu \text{m}$ , respectively, and the alignment precision between the longitudinal beam tunnel and the serpentine waveguide is $\pm 2 \mu \text{m}$ , which are primarily determined by the process parameters. This article explains—in detail—the effects of processing parameters on patterns such as soft bake, exposure dose, post-bake, and electroforming current. The transmission properties of the microfabricated 340-GHz FWG are measured using an Agilent vector network, and the results verify its performance. The capability of making two-layer structures demonstrates that a two-step UV-LIGA technology has great promise for microfabricating FWGs on the order of micrometers for THz vacuum electronic devices.

The traditional rearview mirror method cannot fully guarantee safety when driving trucks. RGB and infrared images collected by cameras are used for registration and recognition, so as to achieve the perception of surroundings and ensure safe driving. The traditional scale-invariant feature transform (SIFT) algorithm has a mismatching rate, and the YOLO algorithm has an optimization space in feature extraction. To address these issues, this paper proposes a truck surround sensing technique based on multi-features and an improved YOLOv5 algorithm. Firstly, the edge corner points and infrared features of the preset target region are extracted, and then a feature point set containing the improved SIFT algorithm is generated for registration. Finally, the YOLOv5 algorithm is improved by fusing infrared features and introducing a composite prediction mechanism at the prediction end. The simulation results show that, on average, the image stitching accuracy is improved by 17%, the time is reduced by 89%, and the target recognition accuracy is improved by 2.86%. The experimental results show that this method can effectively perceive the surroundings of trucks, accurately identify targets, and reduce the missed alarm rate and false alarm rate.

In response to the issue that the fusion process of infrared and visible images is easily affected by lighting factors, in this paper, we propose an adaptive illumination perception fusion mechanism, which was integrated into an infrared and visible image fusion network. Spatial attention mechanisms were applied to both infrared images and visible images for feature extraction. Deep convolutional neural networks were utilized for further feature information extraction. The adaptive illumination perception fusion mechanism is then integrated into the image reconstruction process to reduce the impact of lighting variations in the fused images. A Median Strengthening Channel and Spatial Attention Module (MSCS) was designed to be integrated into the backbone of YOLOv8. In this paper, we used the fusion network to create a dataset named ivifdata for training the target recognition network. The experimental results indicated that the improved YOLOv8 network saw further enhancements of 2.3%, 1.4%, and 8.2% in the Recall, mAP50, and mAP50-95 metrics, respectively. The experiments revealed that the improved YOLOv8 network has advantages in terms of recognition rate and completeness, while also reducing the rates of false negatives and false positives.