The authors present a fully automatic method for camera calibration in expressway scenes and vehicle localisation on curved roads. Our approach does not depend on specific targets or a priori information and adapts to a wide range of variations in road appearances and camera views. We make three main contributions to automatic camera calibration. Firstly, we propose a Deep Calibration Network to estimate the vanishing point and camera extrinsic parameters from an RGB image in an end‐to‐end manner. The vanishing point and camera rotation angles are then used to calculate the camera focal length according to the perspective projection geometry, thus constructing the complete camera matrix. Secondly, we introduce a Multi‐View Camera Calibration Dataset, which contains a total of 32,712 images for 17 expressway scenes. The dataset covers a wide range of camera views and expressway types, which is challenging. Thirdly, we present a novel Curved Reference Line System adapted to the expressway shape to locate vehicles on curved roads accurately. This coordinate system can effectively transform the vehicle coordinates from a Cartesian system to the vehicle's mileage and horizontal distance relative to the lane line. A comprehensive evaluation on Multi‐View Camera Calibration Dataset shows that our technique can automatically calibrate a single camera at about 40fps on an NVIDIA GTX1060 processor and accurately locate vehicles on curved roads, making it suitable for deployment in expressway surveillance systems.

Monocular 3D object detection is rapidly emerging as a key research direction in autonomous driving, owing to its resource efficiency and ease of implementation. However, existing methods face certain limitations in cross-dimensional feature attention mechanisms and multi-order contextual information modeling, which constrain their detection performance in complex scenes. Thus, we propose MonoAMP, an adaptive multi-order perceptual aggregation algorithm for monocular 3D object detection. We first introduce triplet attention to enhance the interaction of cross-dimensional feature attention. Second, we design an adaptive multi-order perceptual aggregation module. It dynamically captures multi-order contextual information and employs an adaptive aggregation strategy to enhance target perception. Finally, we propose an uncertainty-guided adaptive depth ensemble strategy, which models the uncertainty distribution in depth estimation and effectively fuses multiple depth predictions. Experiments demonstrate that MonoAMP significantly enhances performance on the KITTI dataset at the moderate difficulty level, achieving 16.80% ${AP}_{3D}$ and 24.47% ${AP}_{BEV}$. Additionally, the ablation study shows a 3.78% improvement in object detection accuracy over the baseline method. Compared to other advanced methods, MonoAMP demonstrates superior detection capabilities, especially in complex scenarios.

Leveraging thermal infrared imagery to complement RGB spatial information is a key technology in industrial sensing. This technology enables mobile devices to perform scene understanding through RGB-T semantic segmentation. However, existing networks conduct only limited information interaction between modalities and lack specific designs to exploit the thermal aggregation entropy of the thermal modality, resulting in inefficient feature complementarity within bilateral structures. To address these challenges, we propose Wavelet-CNet for RGB-T semantic segmentation. Specifically, we design a Wavelet Cross Fusion Module (WCFM) that applies wavelet transforms to separately extract four types of low- and high-frequency information from RGB and thermal features, which are then fed back into attention mechanisms for dual-modal feature reconstruction. Furthermore, a Cross-Scale Detail Enhancement Module (CSDEM) introduces cross-scale contextual information from the TIR branch into each fusion stage, aligning global localization through contour information from thermal features. Wavelet-CNet achieves competitive mIoU scores of 58.3% and 85.77% on MFNet and PST900, respectively, while ablation studies on MFNet further validate the effectiveness of the proposed WCFM and CSDEM modules.