Extraction of road features from UAV images using a novel level set segmentation approach

ABSTRACT

A novel hybrid technique for road extraction from UAV imagery is presented in this paper. The suggested analysis begins with image segmentation via Trainable Weka Segmentation. This step uses an immense range of image features, such as detectors for edge detection, filters for texture, filters for noise depletion and a membrane finder. Then, a level set method is performed on the segmented images to extract road features. Next, morphological operators are applied on the images for improving extraction precision. Eventually, the road extraction precision is calculated on the basis of manually digitized road layers. Obtained results indicated that the average proportions of completeness, correctness and quality were 93.52%, 85.79% and 81.01%, respectively. Therefore, experimental results validated the superior performance of the proposed hybrid approach in road extraction from UAV images.

KEYWORDS:

• UAV images
• level set
• trainable Weka segmentation
• remote sensing
• road extraction

Acknowledgements

Authors would like to thank three anonymous reviewers for their useful comments which helped us to improve the quality of the manuscript. In addition, the editorial comments by Professor Hyung-Sup Jung are highly appreciated.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Biswajeet Pradhan http://orcid.org/0000-0001-9863-2054

Additional information

Funding

This research is funded by the the Centre for Advanced Modelling and Geospatial Information Systems (CAMGIS), UTS [grant number 321740.2232335], [grant number 321740.2232357].