Search Results (6)

With the growing deployment of autonomous driving and unmanned systems in road environments, efficiently and accurately performing environmental perception and map construction has become a significant challenge for SLAM systems. In this paper, we propose an innovative SLAM framework comprising a frontend tracking network called SGF-net and a backend filtering mechanism, namely Semantic Gaussian Filter. This framework effectively suppresses dynamic objects by integrating feature point detection and semantic segmentation networks, filtering out Gaussian point clouds that degrade mapping quality, thus enhancing system performance in complex outdoor scenarios. The inference speed of SGF-net has been improved by over 23% compared to non-fused networks. Specifically, we introduce SGF-SLAM (Semantic Gaussian Filter SLAM), a dynamic mapping framework that shields dynamic objects undergoing temporal changes through multi-view geometry and semantic segmentation, ensuring both accuracy and stability in mapping results. Compared with existing methods, our approach can efficiently eliminate pedestrians and vehicles on the street, restoring an unobstructed road environment. Furthermore, we present a map update function, which is aimed at updating areas occluded by dynamic objects by using semantic information. Experiments demonstrate that the proposed method significantly enhances the reliability and adaptability of SLAM systems in road environments. Full article

Due to the complexity of outdoor environments, we still face challenges in collecting up-to-date outdoor road network maps because of high data collection costs, resulting in a lack of navigation road network maps in outdoor scenarios. Existing road network extraction methods are mainly divided into trajectory data-based and remote sensing image-based methods. Due to factors such as tree occlusion, methods based on remote sensing images struggle to extract complete road information in outdoor environments. The methods based on trajectory data mainly use vehicle trajectories to extract two-dimensional roads, lacking three-dimensional (3D) road information such as elevation and slope, which are important for navigation path planning in outdoor scenarios. Given this, this paper proposes a hierarchical map construction method for extracting the three-dimensional outdoor pedestrian road network based on crowdsourced trajectory data. This method models the pedestrian road network as a graph composed of pedestrian areas, intersections, and road segments connecting these areas. Three-dimensional roads within and between the intersection areas are generated hierarchically. Experiments and comparative analyses were conducted using real-world outdoor trajectory datasets. Results show that the proposed method has higher accuracy in extracting 3D road information than existing methods. Full article

Object detection is a classic image processing problem. For instance, in autonomous driving applications, targets such as cars and pedestrians are detected in the road scene video. Many image-based object detection methods utilizing hand-crafted features have been proposed. Recently, more research has adopted a deep learning approach. Object detectors rely on useful features, such as the object’s boundary, which are extracted via analyzing the image pixels. However, the images captured, for instance, in an outdoor environment, may be degraded due to bad weather such as haze and fog. One possible remedy is to recover the image radiance through the use of a pre-processing method such as image dehazing. We propose a dehazing model for image enhancement. The framework was based on the conditional generative adversarial network (cGAN). Our proposed model was improved with two modifications. Various image dehazing datasets were employed for comparative analysis. Our proposed model outperformed other hand-crafted and deep learning-based image dehazing methods by 2dB or more in PSNR. Moreover, we utilized the dehazed images for target detection using the object detector YOLO. In the experimentations, images were degraded by two weather conditions—rain and fog. We demonstrated that the objects detected in images enhanced by our proposed dehazing model were significantly improved over those detected in the degraded images. Full article

Pedestrian behavior recognition in the driving environment is an important technology to prevent pedestrian accidents by predicting the next movement. It is necessary to recognize current pedestrian behavior to predict future pedestrian behavior. However, many studies have recognized human visible characteristics such as face, body parts or clothes, but few have recognized pedestrian behavior. It is challenging to recognize pedestrian behavior in the driving environment due to the changes in the camera field of view due to the illumination conditions in outdoor environments and vehicle movement. In this paper, to predict pedestrian behavior, we introduce a position-information added two-stream convolutional neural network (CNN) with multi task learning that is robust to the limited conditions of the outdoor driving environment. The conventional two-stream CNN is the most widely used model for human-action recognition. However, the conventional two-stream CNN based on optical flow has limitations regarding pedestrian behavior recognition in a moving vehicle because of the assumptions of brightness constancy and piecewise smoothness. To solve this problem for a moving vehicle, the binary descriptor dense scale-invariant feature transform (SIFT) flow, a feature-based matching algorithm, is robust in moving-pedestrian behavior recognition, such as walking and standing, in a moving vehicle. However, recognizing cross attributes, such as crossing or not crossing the street, is challenging using the binary descriptor dense SIFT flow because people who cross the road or not act the same walking action, but their location on the image is different. Therefore, pedestrian position information should be added to the conventional binary descriptor dense SIFT flow two-stream CNN. Thus, learning biased toward action attributes is evenly learned across action and cross attributes. In addition, YOLO detection and the Siamese tracker are used instead of the ground-truth boundary box to prove the robustness in the action- and cross-attribute recognition from a moving vehicle. The JAAD and PIE datasets were used for training, and only the JAAD dataset was used as a testing dataset for comparison with other state-of-the-art research on multitask and single-task learning. Full article

Pedestrians are increasingly exposed to slow-onset disasters (SLODs), such as air pollution and increasing temperatures in urban built environments (BEs). Pedestrians also face risks that arise from the combination of the BE features, the effects of SLODs on the microclimate, their own characteristics (e.g., health and ability), and the way they move and behave in indoor and outdoor BE areas. Thus, the effectiveness of sustainable risk-mitigation solutions for the health of the exposed pedestrians should be defined by considering the overlapping of such factors in critical operational scenarios in which such emergency conditions can appear. This work provides an innovative method to define a BE-oriented pedestrian risk index through a dynamic meso-scale approach that considers the daily variation of risk conditions. The method is ensured by a quick-to-apply approach, which also takes advantage of open-source repositories and tools to collect and manage input data, without the need for time-consuming in situ surveys. The resulting risk conditions are represented through meso-scale maps, which highlight the risk differences between BEs by focusing on their open spaces as fundamental parts of the urban road network. The method is applied to a significant case study (in Milan, Italy). The results demonstrate the ability of the approach to identify key input scenarios for risk assessment and mapping. The proposed methodology can: (1) provide insights for simulation activities in critical BE conditions, thanks to the identification of critical daily conditions for each of the factors and for single and multiple risks and (2) support the development of design and regeneration strategies in SLOD-prone urban BEs, as well as the identification of priority areas in the urban BE. Full article

Since widespread applications of wireless sensors networks, low-speed traffic positioning based on the received signal strength indicator (RSSI) from personal devices with WiFi broadcasts has attracted considerable attention. This study presents a new range-based localization method for outdoor pedestrian positioning by using the combination of offline RSSI distance estimation and real-time continuous position fitting, which can achieve high-position accuracy in the urban road environment. At the offline stage, the piecewise polynomial regression model (PPRM) is proposed to formulate the Euclidean distance between the targets and WiFi scanners by replacing the common propagation model (PM). The online stage includes three procedures. Firstly, a constant velocity Kalman filter (CVKF) is developed to smooth the real-time RSSI time series and estimate the target-detector distance. Then, a least squares Taylor series expansion (LS-TSE) is developed to calculate the actual 2-dimensional coordinate with the replacement of existing trilateral localization. Thirdly, a trajectory-based technique of the unscented Kalman filter (UKF) is introduced to smooth estimated positioning points. In tests that used field scenarios from Guangzhou, China, the experiments demonstrate that the combined CVKF and PPRM can achieve the highly accurate distance estimator of <1.98 m error with the probability of 90% or larger, which outperforms the existing propagation model. In addition, the online method can achieve average positioning error of 1.67 m with the much better than classical methods. Full article