High-definition (HD) maps are becoming increasingly important for autonomous vehicles and advanced driver assistance systems (ADAS), as they provide detailed and accurate information about the road environment. Vision localization using HD maps can improve accuracy, but maps can become outdated. Image perception can be useful for HD map change detection. However, achieving robust and high-precision alignment between images and HD maps is challenging in varying environmental conditions. In addition, evaluating alignment in the absence of ground truth data is inconvenient. This article proposes a trajectory interpolation-based method for reconstructing lane markings from images to detect road changes when compared with the HD grid map. Images and HD maps are registered using a distance transform and cross-entropy-based optimization. The article also proposes a metric based on intersection over union for evaluating alignment accuracy. Experiments were conducted using a third-party-collected campus dataset and the Argoverse2 open-source dataset to demonstrate the effectiveness of the proposed methods in detecting road changes and achieving high-precision image and HD map fusion.

Self-supervised monocular scene flow estimation, aiming to understand both 3D structures and 3D motions from two temporally consecutive monocular images, has received increasing attention for its simple and economical sensor setup. However, the accuracy of current methods suffers from the bottleneck of less-efficient network architecture and lack of motion rigidity for regularization. In this paper, we propose a superior model named EMR-MSF by borrowing the advantages of network architecture design under the scope of supervised learning. We further impose explicit and robust geometric constraints with an elaborately constructed ego-motion aggregation module where a rigidity soft mask is proposed to filter out dynamic regions for stable ego-motion estimation using static regions. Moreover, we propose a motion consistency loss along with a mask regularization loss to fully exploit static regions. Several efficient training strategies are integrated including a gradient detachment technique and an enhanced view synthesis process for better performance. Our proposed method outperforms the previous self-supervised works by a large margin and catches up to the performance of supervised methods. On the KITTI scene flow benchmark, our approach improves the SF-all metric of the state-of-the-art self-supervised monocular method by 44% and demonstrates superior performance across sub-tasks including depth and visual odometry, amongst other self-supervised single-task or multi-task methods.

In existing self-supervised depth and ego-motion estimation methods, ego-motion estimation is usually limited to only leveraging RGB information. Recently, several methods have been proposed to further improve the accuracy of self-supervised ego-motion estimation by fusing information from other modalities, e.g., depth, acceleration, and angular velocity. However, they rarely focus on how different fusion strategies affect performance. In this paper, we investigate the effect of different fusion strategies for ego-motion estimation and pro-pose a new framework for self-supervised learning of depth and ego-motion estimation, which performs ego-motion estimation by leveraging RGB and inferred depth information in a Multi-Layer Fusion manner. As a result, we have achieved state-of-the-art performance among learning-based methods on the KITTI odometry benchmark. Detailed studies on the design choices of leveraging inferred depth information and fusion strategies have also been carried out, which clearly demonstrate the advantages of our proposed framework.3

Summary Although the construction of steel sheet pile cofferdam has good practicability in the process of water conservancy project construction, the construction period of the water project is long due to the large amount of work, and the cofferdam itself is greatly affected by the water level, topography, geological period, and other factors. With the continuation of time and the change of complex hydrogeological environment, it is easy to cause the accumulation of hidden safety hazards in the construction of the project during the construction period, and the unreasonable and untimely risk warning and control have led to some major construction accidents. In this paper, the SVM (Support Vector Machine) medium‐ and short‐term water level prediction model is established. The SVM tool is used to establish a prediction model that takes complex hydrological scenes and weather changes into account comprehensively, so that the medium‐ and short‐term water level can be predicted more accurately, thus achieving dynamic adjustment and better adapting to the actual requirements of steel sheet pile cofferdam construction. The results show that there is a good co‐integration relationship between the prediction factors selected by the medium‐ and short‐term water level prediction model, which proves the rationality of the multivariable predictions selected in this paper. At the same time, in the precipitation concentration period, the relative error of the SVM prediction model is relatively small, and it can achieve dynamic water level prediction with the update of the medium‐ and short‐term weather forecast, which can meet the requirements of engineering construction and accuracy.

Catadioptric panoramic vision systems have been widely used in many fields because of their advantages such as a wide field of view, integrated imaging, and rotational symmetry. They also play a very important role in the monitoring of unmanned platforms. However, the resolution of catadioptric panoramic vision systems is not very high, usually less than 5 million pixels at present. Even if the resolution is greater than 5 million pixels, the unwrapping and rectification of these images is carried out off-line. Further, the systems are also deployed in a stationary or near-stationary state. This paper proposes an unwrapping and rectification method for a high-resolution catadioptric panoramic vision system that can be used while moving. It can segment dynamic circular mark regions from panoramic videos accurately, determine the center coordinates of circular images in real time, eliminate the processing of unwanted video regions, and shorten the image processing time. In addition, the center coordinates and radius of the circular mark regions are obtained simultaneously so that the image distortion caused by inaccurate center coordinates is reduced. The method uses radial and decentering distortion parameters and a correction factor for fitting a polynomial to rectify the panoramic video without distortion.