Traffic Lane Detection using Fully Convolutional Neural Network

Zang, Jinju; Zhou, Wei; Zhang, Guanwen; Duan, Zhemin

doi:10.23919/apsipa.2018.8659684

Cited by 18 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, the Lspectra variable, the characteristic spectrum of line markings, was first obtained and used to match line marking points. Alternatively, deep learning algorithms have been employed to replace feature extraction steps, as developed in [14], where Zang et al adopted a fully convolutional neural network (CNN) to achieve the pixel-wise detection of lane lines. The detection accuracy of the best model reached 82.24% on the test dataset, where there were high reflection cases.…”

Section: Related Workmentioning

confidence: 99%

Vision-Based Pavement Marking Detection and Condition Assessment—A Case Study

Wang

et al. 2021

Applied Sciences

View full text Add to dashboard Cite

Pavement markings constitute an effective way of conveying regulations and guidance to drivers. They constitute the most fundamental way to communicate with road users, thus, greatly contributing to ensuring safety and order on roads. However, due to the increasingly extensive traffic demand, pavement markings are subject to a series of deterioration issues (e.g., wear and tear). Markings in poor condition typically manifest as being blurred or even missing in certain places. The need for proper maintenance strategies on roadway markings, such as repainting, can only be determined based on a comprehensive understanding of their as-is worn condition. Given the fact that an efficient, automated and accurate approach to collect such condition information is lacking in practice, this study proposes a vision-based framework for pavement marking detection and condition assessment. A hybrid feature detector and a threshold-based method were used for line marking identification and classification. For each identified line marking, its worn/blurred severity level was then quantified in terms of worn percentage at a pixel level. The damage estimation results were compared to manual measurements for evaluation, indicating that the proposed method is capable of providing indicative knowledge about the as-is condition of pavement markings. This paper demonstrates the promising potential of computer vision in the infrastructure sector, in terms of implementing a wider range of managerial operations for roadway management.

show abstract

Section: Related Workmentioning

confidence: 99%

Vision-Based Pavement Marking Detection and Condition Assessment—A Case Study

Wang

et al. 2021

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Emphasizing the lane markings detection under a distinct traffic scene, Zang et al utilized a fully connected neural network for identifying lane detection and classification. Merely, the detection rate of the algorithm is still low, as it simulated by lane occlusion, reflection, and illumination changes [15]. Tian et al proposed a compositional method of CNN and recurrent neural network (RNN) for detecting and locating lanes as a tiny object in which convolution layers are used instead of up and down sampling.…”

Section: Related Workmentioning

confidence: 99%

An efficient encode-decode deep learning network for lane markings instant segmentation

Mamun

Ping

Hossen

2021

IJECE

View full text Add to dashboard Cite

<span lang="EN-US">Nowadays, advanced driver assistance systems (ADAS) has been incorporated with a distinct type of progressive and essential features. One of the most preliminary and significant features of the ADAS is lane marking detection, which permits the vehicle to keep in a particular road lane itself. It has been detected by utilizing high-specialized, handcrafted features and distinct post-processing approaches lead to less accurate, less efficient, and high computational framework under different environmental conditions. Hence, this research proposed a simple encode-decode deep learning approach under distinguishing environmental effects like different daytime, multiple lanes, different traffic condition, good and medium weather conditions for detecting the lane markings more accurately and efficiently. The proposed model is emphasized on the simple encode-decode Seg-Net framework incorporated with VGG16 architecture that has been trained by using the inequity and cross-entropy losses to obtain more accurate instant segmentation result of lane markings. The framework has been trained and tested on a vast public dataset named Tusimple, which includes around 3.6K training and 2.7 k testing image frames of different environmental conditions. The model has noted the highest accuracy, 96.61%, F1 score 96.34%, precision 98.91%, and recall 93.89%. Also, it has <span>also obtained the lowest 3.125% false positive and 1.259% false-negative value, which transcended some of the previous researches. It is expected to</span> assist significantly in the field of lane markings detection applying deep neural networks.</span>

show abstract

“…It provides a decision basis for the steering and lane change of the vehicle. At present, there are two main methods for lane detection: Deep Learning (DL) [1][2][3] and traditional feature extraction. DL mainly uses a multi-layer convolutional neural network and a large number of samples are used to complete lane line recognition through continuous learning, without the need of manually designed features.…”

Section: Introductionmentioning

confidence: 99%

A Novel Lane Line Detection Based on Multi-feature Fusion and Windows Searching

Cai

Gao

Pan

et al. 2021

Preprint

View full text Add to dashboard Cite

Lane detection embedded in intelligent vehicles can greatly improve the security of automatic driving. This work offers a new approach towards lane detection in the video in real-time combining multi-feature fusion and window searching. As the pre-procession, polygon filling is adopted to locate ROI (Region of Interest) in the video frames, which contain the lane-lines to be detected. To remove the backgrounds in the ROIs, we extract and fuse the features of color, histogram, and gradient of line lanes. Based on the density distribution of the pixels in the line lanes, the initial location is found by homography transformation. Then all candidate pixel points in the whole lane-line are extracted in the way of window searching. Finally, On the basis of the obtained lane-mark coordinates, a curve model is defined, and the model parameters are obtained by Least Square Estimation (LSE). Experimental results show the robustness and instantaneity of the proposed algorithm with the accuracy of 96% and the detecting time of only 20.7ms. In addition, lane-lines with misleading backdrops can also be detected such as yellow lane lines on the ground, shadow, bright light, lane-line defects and traffic light

show abstract

Traffic Lane Detection using Fully Convolutional Neural Network

Cited by 18 publications

References 15 publications

Vision-Based Pavement Marking Detection and Condition Assessment—A Case Study

Vision-Based Pavement Marking Detection and Condition Assessment—A Case Study

An efficient encode-decode deep learning network for lane markings instant segmentation

A Novel Lane Line Detection Based on Multi-feature Fusion and Windows Searching

Contact Info

Product

Resources

About