Autonomous driving, as a pivotal technology in modern transportation, is progressively transforming the modalities of human mobility. In this domain, vehicle detection is a significant research direction that involves the intersection of multiple disciplines, including sensor technology and computer vision. In recent years, many excellent vehicle detection methods have been reported, but few studies have focused on summarizing and analyzing these algorithms. This work provides a comprehensive review of existing vehicle detection algorithms and discusses their practical applications in the field of autonomous driving. First, we provide a brief description of the tasks, evaluation metrics, and datasets for vehicle detection. Second, more than 200 classical and latest vehicle detection algorithms are summarized in detail, including those based on machine vision, LiDAR, millimeter-wave radar, and sensor fusion. Finally, this article discusses the strengths and limitations of different algorithms and sensors, and proposes future trends.

Power line extraction is not only crucial for UAVs obstacle avoidance, but also a fundamental step for fault diagnosis of power lines. Therefore, achieving robust and accurate extraction of power lines in aerial images is essential to enable intelligent UAVs inspection. Unfortunately, power line extraction is an extremely challenging task, and all the current methods attempt to utilize a single model to solve the problem of power line extraction in complex and variable scenes. This results in insufficient generalization ability and suboptimal computational efficiency. In this work, we propose a power line scene classification network based on complexity assessment, named SceneNet, which can provide a solution for tackling power line extraction challenges. Firstly, we propose a human-machine hybrid reasoning model to obtain the ground truth of image complexity reasonably and build the first benchmark dataset that can be used for automatic classification research of power line scenes. Secondly, we propose an improved StyleGAN3 model and loop transfer learning strategy for data augmentation. Most importantly, the SceneNet comprises a multi-feature joint embedding module and a feature encoding-decoding module. On the one hand, it achieves the multi-level fusion of artificial features and high-dimensional semantic features. On the other hand, we use a self-attention mechanism to enable full use of the contextual association between each block of the fusion feature map. The SceneNet has successfully achieved the mapping and pattern recognition between the abstract concept and the concrete features. Experimental results demonstrate that the SceneNet is obviously superior to the existing 12 state-of-the-art models, and it provides guidance and delineation of applicable scenes for power line extraction methods.

The mycotoxin ochratoxin A (OTA) is toxic to humans and frequently contaminates wine and beer. Antibodies are essential recognition probes for the detection of OTA. However, they have several drawbacks, such as high costs and difficulty in preparation. In this study, a novel magnetic-bead-based automated strategy for efficient and low-cost OTA sample preparation was developed. Human serum albumin, which is an economical and stable receptor based on the mycotoxin–albumin interaction, was adapted and validated to replace conventional antibodies to capture OTA in the sample. Ultra-performance liquid chromatography–fluorescence detection was used in combination with this preparation method for efficient detection. The effects of different conditions on this method were investigated. The recovery of OTA samples spiked at three different concentrations ranged from 91.2% to 102.1%, and the relative standard deviations (RSDs) were 1.2%–8.2% in wine and beer. For red wine and beer samples, the LODs were 0.37 and 0.15 µg/L, respectively. This reliable method overcomes the drawbacks of conventional methods and offers significant application prospects.

A chromatography-free detection of aflatoxin B₁ (AFB₁) in cereals and oils through atomic absorption spectroscopy (AAS) has been developed using quantum dots and immunomagnetic beads. A magneto-controlled pretreatment platform for automatic purification, labeling, and digestion was constructed, and AFB₁ detection through AAS was enabled. Under optimal conditions, this immunoassay exhibited high sensitivity for AFB₁ detection, with limits of detection as low as 0.04 μg/kg and a linear dynamic range of 2.5–240 μg/kg. The recoveries for four different food matrices ranged from 92.6% to 108.7%, with intra- and inter-day standard deviations of 0.7–6.3% and 0.6–6.9%, respectively. The method was successfully applied to the detection of AFB₁ in husked rice, maize, and polished rice samples, and the detection results were not significantly different from those of liquid chromatography-tandem mass spectrometry. The proposed method realized the detection of mycotoxins through AAS for the first time. It provides a new route for AFB₁ detection, expands the application scope of AAS, and provides a reference for the simultaneous determination of multiple poisonous compounds (such as mycotoxins and heavy metals).

Robust power line detection from aerial images is an essential step for intelligent unmanned aerial vehicles (UAVs) inspection. However, compared to regular object detection tasks, the generalization ability and accuracy of current methods are seriously insufficient due to the variability of power line scenes and the low proportion of power lines in the image. In this letter, we propose an object-based Markov random field with discrete multi-neighborhood system (OMRF-DMNS) model, which can provide a solution for tackling power line detection challenges. Firstly, the proposed method constructs a discrete multi-neighborhood system (DMNS) for the non-contact nodes. This not only greatly improves the anti-noise ability of the model, but also lays the foundation for the transfer of context information among nodes under different receptive fields. Secondly, a multi-level logical model with multi-neighborhood joint reasoning (MLL-MNJR) is defined on the DMNS, which makes full use of the line features and location information to achieve semantic inference of nodes. Therefore, the OMRF-DMNS model can fully mine the multi-dimensional feature associations and context information among nodes, enabling it to effectively distinguish between power line segments and interference noises. Experimental results demonstrate that the proposed method significantly outperforms previous methods in different power line scenes.

Integrated sensing and communication (ISAC) has emerged as a transformative technology for intelligent transportation systems. Index modulation (IM), recognized for its high robustness and energy efficiency (EE), has been successfully incorporated into ISAC systems. However, most existing IM-based ISAC schemes overlook the spatial multiplexing potential of millimeter-wave channels and remain confined to single-hop vehicle-to-vehicle (V2V) setups, failing to address the challenges of energy consumption and noise accumulation in real-world multi-hop V2V networks with complex road topologies. To bridge this gap, we propose a spatial scattering modulation-based ISAC (ISAC-SSM) scheme and introduce it to multi-hop V2V networks. The proposed scheme leverages the sensed positioning information to select maximum signal-to-noise ratio relay vehicles and employs a detect-amplify-and-forward (DAF) protocol to mitigate noise propagation, while utilizing sensed angle data for Doppler compensation to enhance communication reliability. At each hop, the transmitter modulates index bits on the angular-domain spatial directions of scattering clusters, achieving higher EE. We initially derive a closed-form bit error rate expression and Chernoff upper bound for the proposed DAF ISAC-SSM under multi-hop V2V networks. Both theoretical analyses and Monte Carlo simulations have been made and demonstrate the superiority of DAF ISAC-SSM over existing alternatives in terms of EE and error performance. Specifically, in a two-hop network with 12 scattering clusters, compared with DAF ISAC-conventional spatial multiplexing, DAF ISAC-maximum beamforming, and DAF ISAC-random beamforming, the proposed DAF ISAC-SSM scheme can achieve a coding gain of 1.5 dB, 2 dB, and 4 dB, respectively. Moreover, it shows robust performance with less than a 1.5 dB error degradation under 0.018 Doppler shifts, thereby verifying its superiority in practical vehicular environments.