Search

Lane detection is a fundamental component of advanced driver assistance systems, facilitating critical functionalities like Lane Keep/Change Assistance, Lane Departure Warning, Adaptive Cruise Control, and Vehicle Localization. Despite significant advancements in camera-based lane detection, it continues to confront challenges that can be effectively addressed with LiDAR technology. This research contributes to the domain of LiDAR-based lane detection across three pivotal areas. Firstly, we introduce the BoostedDim Attention method, enhancing traditional Multi-Head Self-Attention (MHA) calculations within the shallow Vision Transformers-based K-Lane baseline model. This method excels, particularly in demanding scenarios, including unknown and nighttime conditions at short ranges (0–30 m) and daytime scenarios for long ranges (30–50 m). Secondly, we devise distance-based True Positive Rate (TPR) and Lateral Error evaluation metrics, providing a more precise and tailored approach to evaluating model performance compared to conventional metrics. These metrics consider sensor-specific and task-specific factors, offering a comprehensive assessment of LiDAR-based lane detection capabilities. Lastly, our investigation sheds light on the significance of calibrated reflectivity and intensity data, revealing their impact on lane detection under various lighting conditions. Notably, we highlight the positive influence of intensity data in low-light conditions for short ranges and its adverse effect during daytime for long ranges. These findings have significant implications for enhancing autonomous driving applications and other computer vision tasks.

Lane detection is a key technical challenge in autonomous driving and advanced driving assistance systems. The current lane detection system has insufficient robustness and low detection accuracy problem in complex and harsh environmental conditions, and lacks research on visual cues without information. In order to accurately and effectively detect multiple lanes in challenging scenarios and meet the real-time requirements of actual driving, a lane detection method based on recurrent cross attention mechanism is proposed in this paper. Firstly, by embedding the recurrent cross attention module in the feature extraction module, the contextual information of surrounding pixels on the intersection path is obtained, and useful information is collected from other regions for accurate lane detection; Secondly, anchor point detection is selected based on the prior position of the designed lane starting position; Finally, combining mixed anchor points to represent lane coordinates, using row and column anchors to learn lane coordinates and improve the real-time performance of lane detection models. The detection model RCCLaneNet designed in this paper was evaluated on the TuSimple and CULane datasets. The experimental results showed that the F1 of this method can reach 97.05% on the TuSimple dataset, 78.5% on the CULane dataset, and the real-time inference speed index FPS can reach +150. The proposed model achieves advanced performance in lane detection tasks in complex scenes while maintaining real-time detection efficiency, and has good universality in actual driving scenarios.

Lanes are different sections of a roadway that are marked or assigned to traffic movement. They are used to organize and govern the passage of vehicles on roads. A lane is essential for the visual navigation system of an autonomous vehicle. The concept of a lane represents a traffic sign that has significant meaning. However, it also exhibits a unique local pattern that requires detailed low-level characteristics to identify its location accurately. Utilizing various feature levels is crucial for achieving effective lane recognition. Therefore, this study used a Cross-Layer Refinement Network (CLRNet) to enhance lane recognition by incorporating high and low-level lane-detecting characteristics. This approach involves identifying lanes based on high-level semantic properties, and refining them using low-level features. The proposed method aims to improve the localization accuracy by leveraging additional contextual information and local-specific lane characteristics. The network architecture combines the elements of LeNet-5 and AlexNet, utilizing the more profound architecture of AlexNet for complex feature learning and localized pattern recognition from LeNet-5. A global context is acquired to enhance the lane feature representation. The Line Intersection over Union (LIoU) loss function, which treats the lane line as a whole unit rather than as individual segments, is employed to enhance the localization accuracy. The experimental results demonstrate the superior performance of the proposed method compared to existing state-of-the-art lane detection algorithms.

Parking lot lane detection is a critical component of automated driving technology, requiring high accuracy, speed, and ease of deployment. In this study, we aimed to develop a comprehensive dataset of parking lane lines and construct an innovative parking lot recognition model. The model, named the Point Angle Instance Network (PAINet), effectively clusters each key point by embedding angle information, thus enhancing the stability and training efficiency. Additionally, we have developed an information collection module to address the issue of fault detection of the guidance arrow in parking lot environments. The performance of the model was tested and evaluated using the created parking lane dataset, and promising results were obtained. The model achieved an F1 score of 85.48%, an FPS of 164, and a GFLOPs of 4.1 in the task of parking lot lane detection in a surround-view situation. These results indicate the accuracy, practicality, and real-time performance of the model, highlighting its potential for use in automated driving systems.

Accurate and timely lane detection is imperative for the seamless operation of autonomous driving systems. In this study, leveraging the gradual variation of lane features within a defined range of width and length, we introduce an enhanced Spatial-Temporal Recurrent Neural Network (SCNN) framework. This framework serves as the cornerstone of an innovative hybrid spatial-temporal model for lane detection, which is tailored to address the prevalent issues of substandard detection performance and insufficient real-time processing in intricate scenarios, such as those involving lane erosion and inconsistent lighting conditions, which often challenge conventional models. With the foundational understanding that lanes manifest as continuous lines, we employ a temporal sequence of lane imagery as the input to our model, thereby ensuring a rich provision of feature information. The model adopts an encoder-decoder structure and integrates a Spatial-Temporal Recurrent Neural Network module for the extraction of interrelated information from the image sequence. The model culminates in the output of the lane detection results for the terminal frame. The proposed lane detection model exhibits a commendable synthesis of accuracy and real-time efficiency, attaining an Accuracy of 97.87%, an $F_{1}$ -score of 0.943, and a FPS of 19.342 on the tvtLANE dataset and an Accuracy of 98.21%, an $F_{1}$ -score of 0.957 on the Tusimple dataset. These metrics signify a superior performance over a majority of the current lane detection methods.

Driver assistance systems and autonomous vehicle navigation have become important topics in vehicular technology. Among all of the functions, lane detection is one of the most important. A variety of approaches have been proposed in this field. While learning-based methods achieve impressive accuracy in detecting complex lane markings under clear daylight conditions, adapting these models to diverse weather conditions remains a challenge. On the other hand, geometric-based approaches require parameter tuning for different scenarios but require fewer computational resources. A self-tuned algorithm with high generalizability across diverse weather conditions is proposed in this paper. The algorithm integrates fuzzy logic-based adaptive functions with edge identification and line detection modules, enabling image adjustments in response to challenging weather conditions. The proposed tracking function utilizes previous detection results to fine-tune the selected Range of Interest (ROI), optimizing both accuracy and processing time. By incorporating these adaptive features into common geometric-based frameworks, the algorithm achieves higher detection rates compared to previous studies during challenging weather conditions. Furthermore, the proposed work exhibits better generalizability and significantly shorter processing time when compared to state-of-the-art learning-based models, as demonstrated through extensive testing on multiple datasets.

Recently, with the development of semiconductors and VLSI (Very Large Scale Integrated Circuit), the technology required for autonomous driving is rapidly developing. One of the technologies that cannot be left out is the lane detection function. Lane recognition requires a lot of data from the camera sensor. As a result, the data size increases, making it difficult to process on a lightweight embedded board. This paper proposes a sliding-based parallel segment image processing method to solve this problem. Most boards in autonomous vehicles are lightweight, so the technique has been designed to reduce computation and power consumption. After fetching the image’s pixel data, grayscale conversion, Gaussian smoothing, Sobel operator, non-maximum suppression, and hysteresis are performed in parallel. Lanes were detected by performing a Hough transform operation on an image for which edge detection was completed in parallel. Due to the nature of parallel processing, it is more effective when image input is continuous and numerous than single image processing. This algorithm is written in C language and VHDL (VHSIC Hardware Description Language) for two parts in the board, DE1-SoC, FPGA (Field Programmable Gate Array) and HPS (Hard Processor System. Due to the use of the C language and VHDL, parallel programming uses 3.1 times less time, twice as much memory and slightly more power than sequential programming. For hardware languages such as Verilog, the computation algorithms have been converted to a fixed point. When comparing HPS and FPGA, the FPGA consumed significantly fewer resources, with 18 times shorter run time, 50 times fewer clock cycles, 3 times less power, and 183 times less energy. This provides a substantial benefit.

As one of the most indispensable means of transportation in modern society, vehicles guarantee our daily commuting and logistics transportation. However, with the increasing number of vehicles, vehicles have also caused increasingly serious traffic safety problems while providing convenience to our lives. One of the most common of these is traffic accidents caused by vehicle yaw due to driver distraction. As a potential solution to this problem, lane departure warning systems (LDWS) focus on detecting and determining whether the vehicle is deviating from the driveway, considered an essential part of autonomous driving technology, and have received significant attention in recent years. A large number of different types of LDWS systems have been developed, especially in recent years, with the development of artificial intelligence technology, many methods based on deep learning and machine vision have been proposed. However, it is well known that due to the complexity of the network structure in deep learning-based object detection algorithms, the operation of such methods relies on a large amount of computing power support. However, due to the limitation of the overall energy supply of the vehicle, it is usually unable to support computing power similar to the laboratory level. Therefore, how to realize efficient lane departure warnings under the condition of limited computing power is a critical problem to be solved. Accordingly, in this paper, we propose a novel lightweight LDWS model. Different from deep learning methods of LDWS, our LDWS model LEHA can achieve high accuracy and efficiency by relying only on simple hardware. The proposed LEHA consists of three modules: the image processing module, the lane detection module, and the lane departure recognition module. The image pre-processing module is applied to pre-process the original road image, which can improve the accuracy and efficiency of the following lane detection module. After obtaining the processed image, lane detection begins to detect and label the lanes. Finally, lane departure recognition is used to calculate the deviation distance and direction to determine whether the warning should be initiated. To evaluate the performance of LEHA, we compare our method with other state-of-the-art LDWS models in terms of detection accuracy and processing time under ideal and non-ideal lane conditions based on the KITTI dataset. The experimental results demonstrate that our LEHA outperforms state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Real-time accurate lane detection is essential task in autonomous driving systems. We propose a new inference model AR-NET to implement the real-time lane detection task. Compared to the pixel-level lane segmentation approach, the Grid classification method is used in AR-NET to reduce computational consumption. Although the deep feature extraction network can extract features effectively, it does not pay attention to the connection between different feature lanes. In order to effectively fuse the extracted features, we propose an interaction network. The lane information obtained from the encoding layer is sent to the interaction network, which uses interaction to fuse local and global features to achieve a balance of local and global attention. Attention is also used to tune target detail features, which learn contextual information by establishing correlations between the features extracted in convolution and the attention channel. Experiments on a publicly available lane detection benchmark dataset demonstrate that our method achieves excellent performance in terms of speed and accuracy. This method achieves 200 + frames per second(FPS) with an accuracy of 96.11% on TuSimple. [ABSTRACT FROM AUTHOR]

Learning models that have been known to multi-task are gaining too much attention in various fields. One such area of work where Multi-tasking models are extracting hopeful outcomes is Autonomous Driving. Panoptic Driving Perception has become an inevitable aspect of autonomous driving systems which demands a very good accuracy and speed. Over a decade, various multi-tasking models have been proposed to simultaneously carry out traffic object detection, drivable area segmentation, and lane detection. Among all of them, YOLOP, YOLOPv, and HybridNets are the 3 models that have accomplished the above tasks using the BDD100K Dataset with high excellent accuracy and efficiency, thus achieving the new State-Of-The-Art (SOTA) performance. This paper focuses mainly on comparing the performance of the above-mentioned models in various weather conditions based on accuracy, robustness, and computational speed. [ABSTRACT FROM AUTHOR]

Lane detection is one of the core technologies of assisted driving and automatic driving. To further enhance the ability of lane feature extraction, a lane line recognition algorithm based on deep learning and recurrent multi-feature information aggregator was proposed. Given the problem of model operation speed, the algorithm took the lane detection problem as a classification problem based on row selection cells. Given the lane information aggregation problem in the image, a novel RMFA(recurrent multi-feature information aggregator) method was proposed. The ResNet (residual neural network) was combined with the method to propose a lane line recognition network ResNet-RMFA fused with contextual and multi-channel information. The performance of the network model was tested on the open datasets of Tusimple and CULane. The experimental results show that the reasoning time of a single frame image of the model can reach 4. 8 ms, the accuracy on the Tusimple dataset is 96. 07%, and the F1 score on the CULane dataset (IoU =0. 5) is 69. 3%, achieving a good balance between speed and accuracy. [ABSTRACT FROM AUTHOR]

Nowadays, most trajectory prediction algorithms have difficulty simulating actual traffic behavior, and there is still a problem of large prediction errors. Therefore, this paper proposes a multi-object trajectory prediction algorithm based on lane information and foresight information. A Hybrid Dilated Convolution module based on the Channel Attention mechanism (CA-HDC) is developed to extract features, which improves the lane feature extraction in complicated environments and solves the problem of poor robustness of the traditional PINet. A lane information fusion module and a trajectory adjustment module based on the foresight information are developed. A socially acceptable trajectory with Generative Adversarial Networks (S-GAN) is developed to reduce the error of the trajectory prediction algorithm. The lane detection accuracy in special scenarios such as crowded, shadow, arrow, crossroad, and night are improved on the CULane dataset. The average F1-measure of the proposed lane detection has been increased by 4.1% compared to the original PINet. The trajectory prediction test based on D2-City indicates that the average displacement error of the proposed trajectory prediction algorithm is reduced by 4.27%, and the final displacement error is reduced by 7.53%. The proposed algorithm can achieve good results in lane detection and multi-object trajectory prediction tasks. [ABSTRACT FROM AUTHOR]

With artificial intelligence technology progressing at a tremendous speed, intelligent driving has got a lot of recognition in recent years. Lane detection is one of the primary functions in self-driving cars. Traditionally, lane detection was done using image processing algorithms and computer vision techniques, which included extraction of areas which are possible lane areas, edge enhancement etc. Deep learning models with new improvements are being introduced till date. Additionally, a self-driving vehicle must be able to recognise traffic signs. In the proposed work a VGG-16 convolutional neural network is used for road segmentation. The model is trained on the KITTI road/lane detection evaluation 2013 dataset. The model performed well with an accuracy of 98.58%. For traffic sign detection, the German traffic sign recognition benchmark dataset is used. A convolutional neural network is used with ADAM optimizer, which gives an accuracy of 95%. [ABSTRACT FROM AUTHOR]

Lane detection under extreme conditions presents a highly challenging task that requires capturing each crucial pixel to predict the complex topology of lane lines and differentiate the various lane types. Existing methods predominantly rely on deep feature extraction networks with substantial parameters or the fusion of multiple prediction modules, resulting in large model sizes, embedding difficulties, and slow detection speeds. This article proposes a Proportional Feature Pyramid Network (PFPN) through fusing the weights into the FPN for lane detection. For obtaining a more accurately detecting result, the cross refinement block is introduced in the PFPN network. The cross refinement block takes the feature maps and anchors as inputs and gradually refines the anchors from high to low level feature maps. In our method, the high-level features are explored to predict lanes coarsely while localdetailed features are leveraged to improve localization accuracy. Extensive experiments on two widely used lane detection datasets, The Chinese Urban Scene Benchmark for Lane Detection (CULane) and the TuSimple Lane Detection Challenge (TuSimple) datasets, demonstrate that the proposed method achieves competitive results compared with several state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

In light of growing metropolitan traffic, safety when driving has grown more and more persuasive. The majority of fatalities on the roads are caused by drivers leaving their lanes while following the required laws. Many of them are a consequence of the driver's disrupted and drowsy attitude. Lane discipline is essential for both driving and protecting pedestrians on the road. The proposed research seeks to detect lane markings in order to produce a safe and advanced traffic system. The proposed device's capabilities include displaying lane line locations to the driver on any exterior display to more complicated programs such as recognizing lane changes within the near future so that injuries caused on the roadways may be avoided. The proposed strategy aims to detect markings for lanes on the roadway by providing a video of the road using computer vision technology to reduce the incidence of accidents. The proposed system was installed in automobiles and cabs to reduce the number of accidents caused by careless driving along the road. In addition, the efficiency of motorists in transportation-related sectors can be tracked in order to detect and report driver irresponsibility and lack of focus on the roads. [ABSTRACT FROM AUTHOR]

As an indispensable part of the vehicle environment perception task, road traffic marking detection plays a vital role in correctly understanding the current traffic situation. However, the existing traffic marking detection algorithms still have some limitations. Taking lane detection as an example, the current detection methods mainly focus on the location information detection of lane lines, and they only judge the overall attribute of each detected lane line instance, thus lacking more fine-grained dynamic detection of lane line attributes. In order to meet the needs of intelligent vehicles for the dynamic attribute detection of lane lines and more perfect road environment information in urban road environment, this paper constructs a fine-grained attribute detection method for lane lines, which uses pixel-level attribute sequence points to describe the complete attribute distribution of lane lines and then matches the detection results of the lane lines. Realizing the attribute judgment of different segment positions of lane instances is called the fine-grained attribute detection of lane lines (Lane-FGA). In addition, in view of the lack of annotation information in the current open-source lane data set, this paper constructs a lane data set with both lane instance information and fine-grained attribute information by combining manual annotation and intelligent annotation. At the same time, a cyclic iterative attribute inference algorithm is designed to solve the difficult problem of lane attribute labeling in areas without visual cues such as occlusion and damage. In the end, the average accuracy of the proposed algorithm reaches 97% on various types of lane attribute detection.

For autonomous driving, it is imperative to perform various high-computation image recognition tasks with high accuracy, utilizing diverse sensors to perceive the surrounding environment. Specifically, cameras are used to perform lane detection, object detection, and segmentation, and, in the absence of lidar, tasks extend to inferring 3D information through depth estimation, 3D object detection, 3D reconstruction, and SLAM. However, accurately processing all these image recognition operations in real-time for autonomous driving under constrained hardware conditions is practically unfeasible. In this study, considering the characteristics of image recognition tasks performed by these sensors and the given hardware conditions, we investigated MTL (multi-task learning), which enables parallel execution of various image recognition tasks to maximize their processing speed, accuracy, and memory efficiency. Particularly, this study analyzes the combinations of image recognition tasks for autonomous driving and proposes the MDO (multi-task decision and optimization) algorithm, consisting of three steps, as a means for optimization. In the initial step, a MTS (multi-task set) is selected to minimize overall latency while meeting minimum accuracy requirements. Subsequently, additional training of the shared backbone and individual subnets is conducted to enhance accuracy with the predefined MTS. Finally, both the shared backbone and each subnet undergo compression while maintaining the already secured accuracy and latency performance. The experimental results indicate that integrated accuracy performance is critically important in the configuration and optimization of MTL, and this integrated accuracy is determined by the ITC (inter-task correlation). The MDO algorithm was designed to consider these characteristics and construct multi-task sets with tasks that exhibit high ITC. Furthermore, the implementation of the proposed MDO algorithm, coupled with additional SSL (semi-supervised learning) based training, resulted in a significant performance enhancement. This advancement manifested as approximately a 12% increase in object detection mAP performance, a 15% improvement in lane detection accuracy, and a 27% reduction in latency, surpassing the results of previous three-task learning techniques like YOLOP and HybridNet. [ABSTRACT FROM AUTHOR]

Lane detection, a crucial component of autonomous driving systems, is in charge of precise lane location to ensure that cars navigate lanes appropriately. However, in challenging conditions like shadows and extreme lighting, lanes may become obstructed or blurred, posing a significant challenge to the lane-detection task as the model struggles to extract sufficient visual information from the image. The current anchor-based lane-detection network detects lanes in complex scenes by mapping anchors to images to extract features and calculating the relationship between each anchor and other anchors for feature fusion. However, it is insufficient for anchors to extract subtle features from images, and there is no guarantee that the information carried by each anchor is valid. Therefore, this study proposes the adaptive cross-scale ROI fusion network (ACSNet) to fully extract the features in the image so that the anchor carries more useful information. ACSNet selects important anchors in an adaptive manner and fuses these important anchors with the original anchors across scales. Through this feature extraction method, the features of different field-of-view ranges under complex road surfaces can be learned, and diversified features can be integrated to ensure that lanes can be well detected under complex road surfaces such as shadows and extreme lighting. Furthermore, due to the slender structure of lane lines, there are relatively few useful features in the images. Therefore, this study also proposes a Three-dimensional Coordinate Attention Mechanism (TDCA) to enhance image features. The Three-dimensional Coordinate Attention Mechanism extensively explores relationships among features in the row, column, and spatial dimensions. It calculates feature weights for each of these dimensions and ultimately performs element-wise multiplication with the entire feature map. Experimental results demonstrate that our network achieves excellent performance on the existing public datasets, CULane and Tusimple. [ABSTRACT FROM AUTHOR]

Self-driving vehicles require the ability to perceive and understand their surroundings, just like human drivers. It entails navigating efficiently on roads, obeying traffic signs and signals, and avoiding collisions with other vehicles and pedestrians. To address the challenges associated with object detection in self-driving cars, an effort was made to demonstrate lane detection using the OpenCV library. To achieve this goal, the well-established probabilistic Hough transform technique is used for line detection. Before applying Hough transforms, several pre-processing techniques are used, including converting the image to grayscale, camera calibration, and implementing a masking filter. In addition, edge detection is performed using the edge detection method. The study also indicates a preference for the use of HSL (Hue, Saturation, and Lightness) and HSV (Hue, Saturation, Value) color spaces. When HSL is applied, white lines appear purer and brighter, resulting in superior performance compared to using HSV specifically to detect white. This algorithm proved particularly effective in detecting straight lanes, which achieved an accuracy ratio of 96.06%. By incorporating these methodologies, the lane detection algorithm implemented with the OpenCV library addresses the challenges of self-driving vehicles, providing them with improved perception capabilities similar to human drivers. [ABSTRACT FROM AUTHOR]

Advanced Driver Assistant System(ADAS) is a critical component of self-driving cars, and lane detection is one of its applications. Convolution Neural Network (CNN) has exhibited its potential to devise state-of-the-art solutions for various computer vision problems and lane detection. However, different scales of lane lines, partial visibility of lane, and the varying resolution of hierarchical features challenge the CNN-based detection methods to localize the lane accurately. In this work, we propose a deep supervision-based model to detect the partially visible lane lines. Hierarchical (Deep) supervision helps the proposed model to trace and classify the lane features at different scales. The simulation of the proposed model gives robust results with partial visibility of lane paint compared to an existing and proven model. The proposed model's accuracy is as high as 92% for partially visible lane and 99% for full visible lane, which stands to the right in its peer work comparison. [ABSTRACT FROM AUTHOR]

Accurately understanding traffic surroundings is crucial for various autonomous and assisted driving scenarios. The visual perception system must capture the entire scene, including vehicle positions, road conditions, and lane configurations. While existing methods co-train models for these tasks simultaneously, they overlook the topological relationships among roads, lanes, and traffic objects in images. In this paper, we propose leveraging inherent structural relations among these tasks to enhance precise panoptic driving perception. We introduce a cross-task relation mining (CRM) method to achieve this goal. Self-attention mechanisms are used to blend key spatial features within each task, and cross-attention facilitates essential information exchange between tasks, resulting in a more comprehensive scene interpretation. Extensive experiments demonstrate the effectiveness of our approach in complex traffic scenarios. • We present a cross-task relation mining technique that enhances the overall comprehension of driving scenarios. • We propose a multi-scale attention-based interaction module to capture mutual priors across all tasks. • Our method improves multi-task consistency and achieves the best performance across all tasks. [ABSTRACT FROM AUTHOR]

Lane detection is a crucial module for today's autonomous driving cars. Detecting road lanes is a challenging task as it varies in color, texture, boundaries and markings. Traditional lane detection techniques detect the lane by applying a model trained with centralized data. As roads vary in urban and rural areas, a more localized and decentralized training technique is desired for accurate and personalized lane detection. Federated learning has recently proved to be a promising technology that trains and prunes the model using local data. Applying federated learning-based lane detection improves the accuracy of detection and also ensures the security and privacy of autonomous cars. This paper proposes FedLANE, a federated learning-based lane detection technique. U-Net, U-Net long short-term memory (LSTM) and AU-Net architectures were explored using a federated learning approach. Experimental analysis using TuSimple and CuLane dataset shows that the FedLANE based lane detection performs similar to that of the traditional deep learning lane detection models. [ABSTRACT FROM AUTHOR]

The task of lane detection has garnered considerable attention in the field of autonomous driving due to its complexity. Lanes can present difficulties for detection, as they can be narrow, fragmented, and often obscured by heavy traffic. However, it has been observed that the lanes have a geometrical structure that resembles a straight line, leading to improved lane detection results when utilizing this characteristic. To address this challenge, we propose a hierarchical Deep Hough Transform (DHT) approach that combines all lane features in an image into the Hough parameter space. Additionally, we refine the point selection method and incorporate a Dynamic Convolution Module to effectively differentiate between lanes in the original image. Our network architecture comprises a backbone network, either a ResNet or Pyramid Vision Transformer, a Feature Pyramid Network as the neck to extract multi-scale features, and a hierarchical DHT-based feature aggregation head to accurately segment each lane. By utilizing the lane features in the Hough parameter space, the network learns dynamic convolution kernel parameters corresponding to each lane, allowing the Dynamic Convolution Module to effectively differentiate between lane features. Subsequently, the lane features are fed into the feature decoder, which predicts the final position of the lane. Our proposed network structure demonstrates improved performance in detecting heavily occluded or worn lane images, as evidenced by our extensive experimental results, which show that our method outperforms or is on par with state-of-the-art techniques. [Display omitted] • Propose a new way to detect lanes by using their straight-line shape prior. • Combine lane features to Hough points by a Deep Hough Transform module. • Incorporating a Dynamic Convolution Module to differentiate between lanes. • Works well even with hard images and is as good as or better than other ways. [ABSTRACT FROM AUTHOR]

Lane departure warning (LDW) is one of the safety innovations in autonomous cars that provides vehicle position monitoring. This technology will alarm if the vehicle moves out of the lane. Lane detection and lane measurement is the main part of LDW. The novelty of this research is the method can measure different lane marks. Very important to know the lane mark that can and cannot be passed. We use semantic segmentation to segment lane mark solid, lane mark dashed, and road. After extracting road and lane marks, we use inverse perspective mapping (IPM) to help calculate the measurement between the car and lane mark. The data is that 374 images were collected from several roads in Makassar City. The model was evaluated using intersection over union (IoU), reaching 79.8% accuracy. The developed system also estimates the measure between the vehicle and lane marks. The lane measurement estimation system’s test results were evaluated using the root mean square error (RMSE) method to reach between 0.025254 and 0.134345. [ABSTRACT FROM AUTHOR]

Lane detection is a vital component of intelligent driving systems, offering indispensable functionality to keep the vehicle within its designated lane, thereby reducing the risk of lane departure. However, the complexity of the traffic environment, coupled with the rapid movement of vehicles, creates many challenges for detection tasks. Current lane detection methods suffer from issues such as low feature extraction capability, poor real-time detection, and inadequate robustness. Addressing these issues, this paper proposes a lane detection algorithm that combines an online re-parameterization ResNet with a hybrid attention mechanism. Firstly, we replaced standard convolution with online re-parameterization convolution, simplifying the convolutional operations during the inference phase and subsequently reducing the detection time. In an effort to enhance the performance of the model, a hybrid attention module is incorporated to enhance the ability to focus on elongated targets. Finally, a row anchor lane detection method is introduced to analyze the existence and location of lane lines row by row in the image and output the predicted lane positions. The experimental outcomes illustrate that the model achieves F1 scores of 96.84% and 75.60% on the publicly available TuSimple and CULane lane datasets, respectively. Moreover, the inference speed reaches a notable 304 frames per second (FPS). The overall performance outperforms other detection models and fulfills the requirements of real-time responsiveness and robustness for lane detection tasks. [ABSTRACT FROM AUTHOR]

Authors of this paper propose a computer vision based autonomous system for the taxiing of an aircraft in the real world. The system integrates both lane detection and collision detection and avoidance models. The lane detection component employs a segmentation model consisting of two parallel architectures. An airport dataset is proposed, and the collision detection model is evaluated with it to avoid collision with any ground vehicle. The lane detection model identifies the aircraft's path and transmits control signals to the steer-control algorithm. The steer-control algorithm, in turn, utilizes a controller to guide the aircraft along the central line with 0.013 cm resolution. To determine the most effective controller, a comparative analysis is conducted, ultimately highlighting the Linear Quadratic Regulator (LQR) as the superior choice, boasting an average deviation of 0.26 cm from the central line. In parallel, the collision detection model is also compared with other state-of-the-art models on the same dataset and proved its superiority. A detailed study is conducted in different lighting conditions to prove the efficacy of the proposed system. It is observed that lane detection and collision avoidance modules achieve true positive rates of 92.59% and 85.19%, respectively. [ABSTRACT FROM AUTHOR]

In order to address the issue of environmental perception in autonomous driving on structured roads, we propose MultiNet-GS, a convolutional neural network model based on an encoder–decoder architecture that tackles multiple tasks simultaneously. We use the main structure of the latest object detection model, the YOLOv8 model, as the encoder structure of our model. We introduce a new dynamic sparse attention mechanism, BiFormer, in the feature extraction part of the model to achieve more flexible computing resource allocation, which can significantly improve the computational efficiency and occupy a small computational overhead. We introduce a lightweight convolution, GSConv, in the feature fusion part of the network, which is used to build the neck part into a new slim-neck structure so as to reduce the computational complexity and inference time of the detector. We also add an additional detector for tiny objects to the conventional three-head detector structure. Finally, we introduce a lane detection method based on guide lines in the lane detection part, which can aggregate the lane feature information into multiple key points, obtain the lane heat map response through conditional convolution, and then describe the lane line through the adaptive decoder, which effectively makes up for the shortcomings of the traditional lane detection method. Our comparative experiments on the BDD100K dataset on the embedded platform NVIDIA Jetson TX2 show that compared with SOTA(YOLOPv2), the mAP@0.5 of the model in traffic object detection reaches 82.1%, which is increased by 2.7%. The accuracy of the model in drivable area detection reaches 93.2%, which is increased by 0.5%. The accuracy of the model in lane detection reaches 85.7%, which is increased by 4.3%. The Params and FLOPs of the model reach 47.5 M and 117.5, which are reduced by 6.6 M and 8.3, respectively. The model achieves 72 FPS, which is increased by 5. Our MultiNet-GS model has the highest detection accuracy among the current mainstream models while maintaining a good detection speed and has certain superiority. [ABSTRACT FROM AUTHOR]

Lane detection is one of the key techniques to realize advanced driving assistance and automatic driving. However, lane detection networks based on deep learning have significant shortcomings. The detection results are often unsatisfactory when there are shadows, degraded lane markings, and vehicle occlusion lanes. Therefore, a continuous multi-frame image sequence lane detection network is proposed. Specifically, the continuous six-frame image sequence is input into the network, in which the scene information of each frame image is extracted by an encoder composed of Swin Transformer blocks and input into the PredRNN. Continuous multi-frame of the driving scene is modeled as time-series by ST-LSTM blocks, and then, the shape changes and motion trajectory in the spatiotemporal sequence are effectively modeled. Finally, through the decoder composed of Swin Transformer blocks, the features are obtained and reconstructed to complete the detection task. Extensive experiments on two large-scale datasets demonstrate that the proposed method outperforms the competing methods in lane detection, especially in handling difficult situations. Experiments are carried out based on the TuSimple dataset. The results show: for easy scenes, the validation accuracy is 97.46%, the test accuracy is 97.37%, and the precision is 0.865. For complex scenes, the validation accuracy is 97.38%, the test accuracy is 97.29%, and the precision is 0.859. The running time is 4.4 ms. Experiments are carried out based on the CULane dataset. The results show that, for easy scenes, the validation accuracy is 97.03%, the test accuracy is 96.84%, and the precision is 0.837. For complex scenes, the validation accuracy is 96.18%, the test accuracy is 95.92%, and the precision is 0.829. The running time is 6.5 ms. [ABSTRACT FROM AUTHOR]

Driver assistance systems have started becoming a key differentiator in automotive space and all major automotive manufacturers have such systems with various capabilities and stages of implementation. The main building blocks of such systems are similar in nature and one of the major building blocks is road lane detection. Even though lane detection technology has been around for decades, it is still an ongoing area of research and there are still several improvements and optimizations that are possible. This paper offers an Optimized Dynamic Origin Technique (Optimized DOT) for lane detection. The proposed optimization algorithm of optimized DOT gives better results in performance and accuracy compared to other methods of lane detection. Analysis of proposed optimized DOT with various edge detection techniques, various threshold levels, various sample dataset and various lane detection methods were done and the results are discussed in this paper. The proposed optimized DOT lane detection average processing time increases by 9.21 % when compared to previous Dynamic Origin Technique (DOT) and 59.09 % compared to traditional hough transform.

Decision-level information fusion methods using radar and vision usually suffer from low target matching success rates and imprecise multi-target detection accuracy. Therefore, a robust target detection algorithm based on the fusion of frequency-modulated continuous wave (FMCW) radar and a monocular camera is proposed to address these issues in this paper. Firstly, a lane detection algorithm is used to process the image to obtain lane information. Then, two-dimensional fast Fourier transform (2D-FFT), constant false alarm rate (CFAR), and density-based spatial clustering of applications with noise (DBSCAN) are used to process the radar data. Furthermore, the YOLOv5 algorithm is used to process the image. In addition, the lane lines are utilized to filter out the interference targets from outside lanes. Finally, multi-sensor information fusion is performed for targets in the same lane. Experiments show that the balanced score of the proposed algorithm can reach 0.98, which indicates that it has low false and missed detections. Additionally, the balanced score is almost unchanged in different environments, proving that the algorithm is robust.