Road detection based on simultaneous deep learning approaches

Almeida, Tiago; Lourenço, Bernardo; Santos, Vítor

doi:10.1016/j.robot.2020.103605

Cited by 21 publications

(17 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several methods have been proposed to improve segmentation results using road boundary features. In [31], Almeida et al combined the results from two separately-trained models: ENet [44] for lane segmentation and LaneNet [45] for lane boundary detection. A post-processing step is applied to create final segmentation masks from the detected lane boundaries by the LaneNet.…”

Section: A Road Detection Methodsmentioning

confidence: 99%

“…In assistive navigation, these errors are more severe than false negatives because they could endanger the blind user. This limitation can be addressed by using post-processing steps (e.g., combining segmentation results from two separatelytrained models [31]) to remove false positives. However, these heuristic steps are less robust than machine learning methods and may fail in complex scenes.…”

Section: Future Directionsmentioning

confidence: 99%

See 1 more Smart Citation

Pedestrian Lane Detection for Assistive Navigation of Vision-Impaired People: Survey and Experimental Evaluation

et al. 2022

View full text Add to dashboard Cite

Pedestrian lane detection is one of the most crucial tasks in assistive navigation for visionimpaired people. It can provide information on walkable regions, help blind people stay on pedestrian lanes, and assist with obstacle detection. An accurate and real-time lane detection algorithm can improve travel safety and efficiency for the visually impaired. However, despite being an important task for assistive navigation systems, pedestrian lane detection in unstructured scenes has not attracted sufficient attention in the research community. Hence, this paper aims to provide a comprehensive review and an experimental evaluation of methods that can be applied for pedestrian lane detection, thereby laying a foundation for future researchers in this area. This study includes methods proposed for pedestrian lane detection, general road detection, and general semantic segmentation. We review these methods in two categories: traditional methods and deep learning methods. We perform an experimental evaluation of representative methods on a large benchmark dataset specifically created for pedestrian lane detection. We hope this paper can serve as an informative guide for researchers in the field of assistive navigation, and facilitate urgently-needed research for blind people.

show abstract

Section: A Road Detection Methodsmentioning

confidence: 99%

Section: Future Directionsmentioning

confidence: 99%

Pedestrian Lane Detection for Assistive Navigation of Vision-Impaired People: Survey and Experimental Evaluation

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The proposed architecture combined of deconvolution and up sampling operations. Almeida et al [23] presented a road detection method comprising of outputs of the deep learning semantic segmentations models i.e. ENet for the structured and unstructured roads.…”

Section: 2) Semantic Segmentation Based Road Detectionmentioning

confidence: 99%

Pixel Level Segmentation Based Drivable Road Region Detection and Steering Angle Estimation Method for Autonomous Driving on Unstructured Roads

Rasib¹,

Butt²,

Riaz³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

With the recent emergence of deep learning, computer vision-based applications have demonstrated better applicability in accomplishing driving tasks including drivable road region detection, lane keeping and steering control in self-driving vehicles. Till recently, numerous lane-marking detection based steering control and lane keeping methods have been proposed to perform autonomous driving on urban well-structured roads. But the matter of fact is that these methods are not feasible on roads where lane markings are not available or faded over time which makes drivable road region detection a crucial task. Also, it is highly difficult task to estimate the steering angle on such highly deteriorated roads using existing road detection and steering angle estimation methods. To the best of our knowledge, there is no standard benchmark available for drivable road region detection and steering angle estimation on unstructured roads. To this end, we present a large-scale dataset for drivable road detection, comprising of 15,000-pixel level high-quality fine annotations. Also, we present a novel end-to-end drivable road region detection and steering angle estimation method to ensure the autonomous driving on unstructured roads. The proposed method performs pixel-level segmentation to extract drivable road region and calculates the lane interception to estimate the steering angle of self-driving vehicles. Lastly, a comprehensive qualitative and quantitative analysis has been carried out to demonstrate the effectiveness of our proposed dataset, road detection and steering angle estimation method. Our benchmark is available at https://carl-dataset.github.io/index/.

show abstract

“…Methods for improving detection accuracy [15] and computing efficiency [16] have been proposed. In addition, a method for recognizing roads in various environments (structured, unstructured, lane/ line-based, and curb) has been proposed [17]. However, the mentioned studies are difficult to be applied to the environments with intersections.…”

Section: Introductionmentioning

confidence: 99%

Vision-Based Branch Road Detection for Intersection Navigation in Unstructured Environment Using Multi-Task Network

Ahn

Lee

Kim

et al. 2022

Journal of Advanced Transportation

View full text Add to dashboard Cite

Autonomous vehicles need a driving method to be less dependent on localization data to navigate intersections in unstructured environments because these data may not be accurate in such environments. Methods of distinguishing branch roads existing at intersections using vision and applying them to intersection navigation have been studied. Model-based detection methods recognize patterns of the branch roads, but are sensitive to sensor noise and difficult to apply to various complex situations. Therefore, this study proposes a method for detecting the branch roads at the intersection using deep learning. The proposed multi-task deep neural network distinguishes the branch road into a shape of rotated bounding boxes and also recognizes the drivable area to consider obstacles inside the box. Through the output of the network, an occupancy grid map consisting of one branch road at an intersection is obtained, which can be used as an input to the existing motion-planning algorithms that do not consider intersections. Experiments in real environments show that the proposed method detected the branch roads more accurately than the model-based detection method, and the vehicle drove safely at the intersection.

show abstract

Road detection based on simultaneous deep learning approaches

Cited by 21 publications

References 20 publications

Pedestrian Lane Detection for Assistive Navigation of Vision-Impaired People: Survey and Experimental Evaluation

Pedestrian Lane Detection for Assistive Navigation of Vision-Impaired People: Survey and Experimental Evaluation

Pixel Level Segmentation Based Drivable Road Region Detection and Steering Angle Estimation Method for Autonomous Driving on Unstructured Roads

Vision-Based Branch Road Detection for Intersection Navigation in Unstructured Environment Using Multi-Task Network

Contact Info

Product

Resources

About