Robust lane detection &amp;amp; tracking based on novel feature extraction and lane categorization

Özgünalp, Umar; Dahnoun, Naim

doi:10.1109/icassp.2014.6855185

Cited by 30 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then the properties of a lane and a vein will be compared and modifications to the SLT algorithm will be proposed for vein detection. detectors [22], the Otsu algorithm [30], Gabor filters [31], local threshold [28], matched filters [24], and the SLT algorithm [32]. [28] concluded that it is the most robust feature extractor among those tested.…”

Section: Features Of a Vein On A Digital Fundus Imagementioning

confidence: 99%

Retinal vessel segmentation using modified symmetrical local threshold

Özgünalp¹

2020

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

Retinal vessel segmentation is important for the identification of many diseases including glaucoma, hypertensive retinopathy, diabetes, and hypertension. Moreover, retinal vessel diameter is associated with cardiovascular mortality. Accurate detection of blood vessels improves the detection of exudates in color fundus images, as well as detection of the retinal nerve, optic disc, or fovea. A retinal vessel is a darker stripe on a lighter background. Thus, the objective is very similar to the lane detection task for intelligent vehicles. A lane on a road is a light stripe on a darker background (i.e. asphalt). For lane detection, the symmetrical local threshold (SLT) is found to be the most robust feature extractor among the tested algorithms in the road marking (ROMA) dataset. Unfortunately, the SLT cannot be applied directly for retinal vessel segmentation. The SLT is a 1D filter and is designed for detecting vertical or close to vertical light stripes with predictable width. In this paper, the SLT is modified to detect dark stripes and four kernels, instead of one, are designed to detect both vertical and horizontal features of a retinal vessel with variable thickness. The proposed algorithm is tested using the High Resolution Fundus (HRF) image database and the accuracy is estimated to be 95.53%. Furthermore, when tested with the Digital Retinal Images for Vessel Segmentation (DRIVE) database, the accuracy is estimated to be 93.69% .

show abstract

Section: Features Of a Vein On A Digital Fundus Imagementioning

confidence: 99%

Retinal vessel segmentation using modified symmetrical local threshold

Özgünalp¹

2020

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

show abstract

“…Others presented a vigorous lane location and an algorithm to adapt to complex situations and to diminish the impact of threshold. Symmetrical local threshold (SLT) is proposed to improve the feature map and for feature extraction (Ozgunalp et al, 2014).…”

Section: Literature Surveymentioning

confidence: 99%

Recent Developments in Lane Departure Warning System: An Analysis

Panda¹,

Mohanty²

2020

Ethics and Information Technology

View full text Add to dashboard Cite

“…For instance, the accuracy of VP estimation will be greatly reduced on complex roadways. Therefore, stereo vision technologies [14], [15] were introduced to address this problem. Ozgunalp et al [15] proposed an automatic VP estimation pipeline with a twostep process: firstly, the disparity is estimated from stereo images, and secondly, a corresponding disparity accumulator is computed.…”

Section: A Lane Detectionmentioning

confidence: 99%

Using DP Towards A Shortest Path Problem-Related Application

Jiao

Fan

et al. 2019

2019 International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

The detection of curved lanes is still challenging for autonomous driving systems. Although current cutting-edge approaches have performed well in real applications, most of them are based on strict model assumptions. Similar to other visual recognition tasks, lane detection can be formulated as a two-dimensional graph searching problem, which can be solved by finding several optimal paths along with line segments and boundaries. In this paper, we present a directed graph model, in which dynamic programming is used to deal with a specific shortest path problem. This model is particularly suitable to represent objects with long continuous shape structure, e.g., lanes and roads. We apply the designed model and proposed an algorithm for detecting lanes by formulating it as the shortest path problem. To evaluate the performance of our proposed algorithm, we tested five sequences (including 1573 frames) from the KITTI database. The results showed that our method achieves an average successful detection precision of 97.5%.

show abstract

Robust lane detection & tracking based on novel feature extraction and lane categorization

Cited by 30 publications

References 12 publications

Retinal vessel segmentation using modified symmetrical local threshold

Retinal vessel segmentation using modified symmetrical local threshold

Recent Developments in Lane Departure Warning System: An Analysis

Using DP Towards A Shortest Path Problem-Related Application

Contact Info

Product

Resources

About