Search Results

Positioning system is a key module of autonomous driving. As for LiDAR SLAM system, it faces great challenges in scenarios where there are repetitive and sparse features. Without loop closure or measurements from other sensors, odometry match errors or accumulated errors cannot be corrected. This paper proposes a construction method of LiDAR anchor constraints to improve the robustness of the SLAM system in the above challenging environment. We propose a robust anchor extraction method that adaptively extracts suitable cylindrical anchors in the environment, such as tree trunks, light poles, etc. Skewed tree trunks are detected by feature differences between laser lines. Boundary points on cylinders are removed to avoid misleading. After the appropriate anchors are detected, a factor graph-based anchor constraint construction method is designed. Where direct scans are made to anchor, direct constraints are constructed.

Image corruptions due to noise, blur, contrast change, etc., could lead to a significant performance decline of Deep Neural Networks (DNN), which poses a potential threat to DNN-based autonomous vehicles. Previous works attempted to explain corruption from a Fourier perspective. By comparing the absolute Fourier spectrum difference between corrupted images and clean images in the RGB color space, they regard the noise from some corruptions (Gaussian noise, defocus blur, etc.) as concentrating on the high-frequency components while others (contrast, fog, etc.) concentrate on the low-frequency components. In this work, we present a new perspective that unifies corruptions as noise from high frequency and thus propose an image augmentation algorithm to achieve a more robust performance against common corruptions. First, we notice the 1/fα statistical rule of the natural image's spectrum and the channels-wise differential sensitivity on the YCbCr color space of the Human Visual System.

LiDAR and camera fusion have emerged as a promising approach for improving place recognition in robotics and autonomous vehicles. However, most existing approaches often treat sensors separately, overlooking the potential benefits of correlation between them. In this paper, we propose a Cross- Modality Module (CMM) to leverage the potential correlation of LiDAR and camera features for place recognition. Besides, to fully exploit potential of each modality, we propose a Local-Global Fusion Module to supplement global coarse-grained features with local fine-grained features. The experiment results on public datasets demonstrate that our approach effectively improves the average recall by 2.3%, reaching 98.7%, compared with simply stacking of LiDAR and camera.

A specially designed generator has been developed to produce poly-disperse droplet streams: A liquid fuel (n-heptane) is metered to an ultrasonic atomizer to produce droplets, which are then carried and accelerated vertically upwards through a nozzle tube by carrier-air flow. Conditions of the streams at the nozzle exit are modulated by varying the length of nozzle tubes, the fuel and carrier-air flow rate. Optical measurement techniques such as direct photography method, schlieren photography and particle image velocimetry (PIV) are employed to characterize its performance characteristics. Effects of the nozzle tube length, the carrier-air and fuel flow rate are investigated to evaluate the performance of the generator. Longer nozzle tubes provide a better flow guidance for the carrier-air, and tend to generate streams with less and smaller droplets due to the transporting losses.

The port structure consisting of spur pile, vertical pile and beam is subjected to impact loads, so its internal stress state of each point will rapidly change over time. Dynamic photoelastic method is used to study the dynamic stress and stress wave propagation. With epoxy resin and other materials, a photoelastic model of beam to column connection structure is processed and product. The dynamic response of the model under the impact load by the free fall is researched by the dynamic photoelastic method, and recorded by the new digital dynamic photoelastic system with a laser source and high-speed photography system. The internal dynamic stress propagation and distribution, the maximum shear stress and the dynamic stress concentration problems can be obtained by analyzing the dynamic response.

As one of the pollutants that cannot be ignored, soot has a great impact on human health, environment, and energy conversion. In this investigation, the effect of residence time (25ms, 35ms, and 45ms) and ammonia on morphology and nanostructure of soot in laminar ethylene flames has been studied under atmospheric conditions and different flame heights (15 mm and 30 mm). The transmission electron microscopy (TEM) and high-resolution transmission electron microscope (HRTEM) are used to obtain morphology of aggregates and nanostructure of primary particles, respectively. In addition, to analyze the nanostructure of the particles, an analysis program is built based on MATLAB software, which is able to obtain the fringe separation distance, fringe length, and fringe tortuosity parameters of primary particles, and has been verified by the multilayer graphene interlayer distance.

This paper reports an experimental investigation of the wake flow behind a 1/12 scale notchback MIRA model at Re = UL/ν = 6.9×105 (where U is free-stream velocity, L the length of the model and ν viscosity). Focus is placed on the flow asymmetry over the backlight and decklid. Forty pressure taps are used to map the surface pressure distribution on the backlight and decklid, while the wake topology is investigated by means of 2D Particle Image Velocimetry. The analysis of the instantaneous pressure signals over the notch configuration clearly shows that the pressure presents a bi-stable behavior in the spanwise direction, characterized by the switches between two preferred values, which is not found in the vertical direction.

In this study, effects of altitude on free diesel spray morphology, macroscopic spray characteristics and air-fuel mixing process were investigated. The diesel spray visualization experiment using high-speed photography was performed in a constant volume chamber which reproduced the injection diesel-like thermodynamic conditions of a heavy-duty turbocharged diesel engine operating at sea level and 1000 m, 2000 m, 3000 m and 4500 m above sea level. The results showed that the spray morphology became narrower and longer at higher altitude, and small vortex-like structures were observed on the downstream spray periphery. Spray penetration increased and spray angle decreased with increasing altitude. At altitudes of 0 m, 1000 m, 2000 m, 3000 m and 4500 m, the spray penetration at 1.45 ms after start of injection (ASOI) were 79.54 mm, 80.51 mm, 81.49 mm, 83.29 mm and 88.92 mm respectively, and the spray angle were 10.9°, 10.8°, 10.7°, 10.4°and 9.8° respectively.

For sensing system, the trustworthiness of the variant sensors is the crucial point when dealing with advanced driving assistant system application. In this paper, an approach to a hybrid camera-radar application of vehicle tracking is presented, able to meet the requirement of such demand. Most of the time, different types of commercial sensors available nowadays specialize in different situations, such as the ability of offering a wealth of detailed information about the scene for the camera or the powerful resistance to the severe weather for the millimeter-wave (MMW) radar. The detection and tracking in different sensors are usually independent. Thus, the work here that combines the variant information provided by different sensors is indispensable and worthwhile. For the real-time requirement of merging the measurement of automotive MMW radar in high speed, this paper first proposes a fast vehicle tracking algorithm based on image perceptual hash encoding.

Gas diffusion layer (GDL), as a critical constituent of the proton exchange membrane fuel cell (PEMFC), plays a key role in mass, heat, electron, and species transport. GDL generally has two distinct layers: a macro-porous substrate (MPS) and a micro-porous layer (MPL). The fibers in MPS and the cracks formed during the deposition process on the surface of MPL change the overall transport capacity and effect the output performance of PEMFC. In this paper, methods of identifying the structural features of fibers and cracks in GDL images based on artificial intelligence are proposed. The block probabilistic Hough transform and the quadric voting based on the weighted K-means algorithm are programmed to realize the fiber feature extraction, and the crack feature extraction is realized by the regional connectivity algorithm and the geometric feature calculation based on the circumscribed graph of the crack region.

In order to research the effect of process parameters (laser power, welding speed, wire-feed speed, spot diameter) on mechanical properties of Zn-coated Steel Laser Brazing Lap Joint for vehicle, the influence of welding parameters on energy input of brazing seam cross section was theoretically analyzed, and then a great number of laser brazing experiments of 0.7mm galvanized steel was carried out. After that, the tensile strength and micro-hardness tests were also done for well-formed joints of galvanized steel formed in the laser brazing. The results show that joints with good mechanical properties and surface morphology can be formed when laser power is in the range of 2500-3200W and the other parameters keep in a specified range. Joint performance significantly reduces when the value of wire-feed speed exceeds 3.0m/min for that a wider brazing seam cross section can’t be formed.

A Joint Probabilistic Data Association (JPDA) multi-objective tracking improvement algorithm based on camera-radar fusion is proposed to address the problems of poor single-sensor tracking performance, unknown target detection probability, and missing valid targets in complex traffic scenarios. First, according to the correlation rule between the target track and the measurement, the correlation probability between the target and the measurement is obtained; then the measurement collection is divided into camera-radar measurement matched target, camera-only measurement matched target, radar-only measurement matched target, and no-match target; and the correlation probability is corrected with different confidence levels to avoid the use of unknown detection probability.

In the Kmeans cluster segmentation used in traffic scenes, there are often zone optimization and over-segmentation problems caused by the algorithm randomly assigning the initial cluster center. In order to improve the target extraction effect in traffic road scenes, this article proposes an improved Kmeans (IM-Kmeans) method. Firstly, search for the histogram peaks of the whole pixels based on hue, saturation, value (HSV) image, and find the initial cluster centers’ positions and number. Secondly, the noise points which are far away from the center pixel are removed, and then the pixels are classified into the nearest cluster center according to its value. Finally, after the clustering model reaches convergence, the area-clustering method is used for another classification to solve the over-segmentation problem. The simulation and experimental comparisons show that the IM-Kmeans algorithm has higher clustering accuracy than the traditional Kmeans algorithm.

Driving posture study is essential for the evaluation of the occupant packaging. This paper presents a method of reconstructing driver’s postures in a real vehicle using a 3D laser scanner and Human Builder (HB), the digital human modeling tool under CATIA. The scanning data was at first converted into the format readable by CATIA, and then a personalized HB manikin was generated mainly using stature, sitting height and weight. Its pelvis position and joint angles were manually adjusted so as to match the manikin with the scan envelop. If needed, a fine adjustment of some anthropometric dimensions was also preceded. Finally the personalized manikin was put in the vehicle coordinate system, and joint angels and joint positions were extracted for further analysis.

Environmental perception is a crucial subsystem in autonomous vehicles. In order to build safe and efficient traffic transportation, several researches have been proposed to build accurate, robust and real-time perception systems. Camera and LiDAR are widely equipped on autonomous self-driving cars and developed with many algorithms in recent years. The fusion system of camera and LiDAR provides state-of the-art methods for environmental perception due to the defects of single vehicular sensor. Extrinsic parameter calibration is able to align the coordinate systems of sensors and has been drawing enormous attention. However, differ from spatial alignment of two sensors’ data, joint calibration of multi-sensors (more than two sensors) should balance the degree of alignment between each two sensors.

Lane marking detection plays a crucial role in Autonomous Driving Systems or Advanced Driving Assistance System. Vision based lane marking detection technology has been well discussed and put into practical application. LiDAR is more stable for challenging environment compared to cameras, and with the development of LiDAR technology, price and lifetime are no longer an issue. We propose a lane marking detection algorithm based on solid-state LiDARs. First a series of data pre-processing operations were done for the solid-state LiDARs with small field of view, and the needed ground points are extracted by the RANSAC method. Then, based on the OTSU method, we propose an approach for extracting lane marking points using intensity information.

The accuracy of parking slot detection is a challenge for the safety of the Automated Valet Parking (AVP), while traditional methods of range sensor-based parking slot detection have mostly focused on the detection rate in a scenario, where the ego-vehicle must pass by the slot. This paper uses three-dimensional Light Detection And Ranging (3D LiDAR) to efficiently search parking slots around without passing by them and highlights the accuracy of detecting and tracking. For this purpose, a universal process of 3D LiDAR-based high-accuracy slot perception is proposed in this paper. First, the method Minimum Spanning Tree (MST) is applied to sort obstacles, and Separating Axis Theorem (SAT) are applied to the bounding boxes of obstacles in the bird’s-eye view, to find a free space between two adjacent obstacles. These bounding boxes are obtained by using common point cloud processing methods.

A formal gesture recognition based on optics has limitations, but millimeter-wave (MMW) radar has shown significant advantages in gesture recognition. Therefore, the MMW radar has become the most promising human-computer interaction equipment, which can be used for human-computer interaction of vehicle personnel. This paper proposes a multi-branch network based on a residual neural network (ResNet) to solve the problems of insufficient feature extraction and fusion of the MMW radar and immense algorithm complexity. By constructing the gesture sample library of six gestures, the MMW radar signal is processed and coupled to establish the relationship between the motion parameters of the distance, speed, and angle of the gesture information and time, and the depth features are extracted. Then the three depth features are fused. Finally, the classification and recognition of MMW radar gesture signals are realized through the full connection layer.

Automotive embedded DCS is widely used to solve automotive control problems. This paper presents a model-driven development technology for such systems. Models of automotive embedded DCS are built up strictly complying with the four-layer-model architecture, which is presented by Model-Driven Architecture (MDA). Three kinds of models are used to describe the protocol data structure, the algorithm process and visualization aspects of automotive embedded DCS. Corresponding XML databases are created based upon these models. As a single data source, these databases play key roles in further development phases, including generating the protocol specification, MC&D systems and embedded programming, etc. Some demonstrative applications are presented in this paper.

The main challenge for future autonomous vehicles is to identify their location and body pose in real time during driving, that is, “where am I? and how will I go?”. We address the problems of pose estimation and scene perception from continuous visual frames in intelligent vehicle. Recent advanced technology in the domain of deep learning proposes to train some learning models for vehicle’s series detection tasks in a supervised or unsupervised manner, which has numerous advances over traditional approaches, mainly reflected in the absence of manual calibration and synchronization of the camera and IMU. In the paper, we propose a novel approach for pose estimation and scene recognition with a deep fusion of multi-modal neural features in the manner of unsupervised. Firstly, low-cost camera and IMU are used to extract original visual and inertial data, then the visual and inertial encoders are utilized to encoder the feature of the two modes.