Investigation of landslide-based road surface deformation in mountainous areas with single period UAV data

Zeybek, Mustafa; Biçici, Serkan

doi:10.1080/10106049.2022.2142969

Cited by 3 publications

(4 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The methodology effectively identified both longitudinal and transverse cracks on the road surface. Zeybek and Biçici [35] proposed a methodology for surveying road deformation caused by landslides using data collected from a single UAV flight rather than multiple flights over time. It found that there are areas of road subsidence and elevation changes of up to 20 cm.…”

Section: Literature Reviewmentioning

confidence: 99%

“…The DJI Phantom 4 RTK has been used in numerous studies before, and its features have been summarized in these studies [5,36]. Additionally, the data collection parameters used in this study have also been explained in detail in previous studies [5,35].…”

Section: Geographical Setupmentioning

confidence: 99%

“…During the flight, 165 images were collected from study area 1, and 116 images were collected from study area 2. These raw images were then processed using the commercial Structure From Motion (SfM) based software called Pix4D Mapper to generate a 3D point cloud [5,35]. Figures 2(a) and (b) show the point clouds produced for the study area 1 and 2 on the orthomosaic, respectively.…”

Section: Geographical Setupmentioning

confidence: 99%

“…Figures 2(a) and (b) show the point clouds produced for the study area 1 and 2 on the orthomosaic, respectively. 3D point cloud generation and the parameters used have been described in detail in previous studies [5,35].…”

Section: Geographical Setupmentioning

confidence: 99%

See 3 more Smart Citations

Different KNN parameters on road surface extraction from UAV-based point cloud under several machine learning algorithms

Biçici

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Accurate road surface from a three-dimensional (3D) point cloud depends on various parameters. One crucial parameter is the set of point features. Point features enable classification by capturing characteristics of the surface on which the points are located. These features are calculated based on the closest points surrounding each point. In this study, the K-nearest neighbors algorithm (KNN) was applied to identify these closest points. The KNN algorithm requires only one input, the number of closest points (k). Eight different point features were developed using different k values, and their impact on road surface classification from the 3D point cloud was investigated. It was observed that there is no significant improvement in classification accuracy until a certain k value. However, better classification accuracy was achieved after a certain k value. The effect of different k values was also investigated under different training sample structures and machine learning (ML) algorithms. When training samples were selected from a single location as a large group, similar classification accuracy was obtained across different k values. Conversely, when training samples were chosen from various regions in smaller groups rather than a single large group, improved classification was observed as the k value increased. Additionally, it was noted that five different ML algorithms-random forest, support vector machine, generalized linear model, linear discriminant analysis, and robust linear discriminant analysis-have almost similar performance under different k values. Finally, using the optimum k value, improvements of up to 4.543% and 6.601% in accuracy and quality measures, respectively, were found.

show abstract

Section: Literature Reviewmentioning

confidence: 99%