Abstract
The ability to classify asphalt surfaces is an important goal for the selection of suitable non-variant targets as pseudo-invariant targets during the calibration/validation of remotely-sensed images. In addition, the possibility to recognize different types of asphalt surfaces on the images can help optimize road network management. This paper presents a multi-resolution study to improve asphalt surface differentiation using field spectroradiometric data, laboratory analysis and remote sensing imagery. Multispectral Infrared and Visible Imaging Spectrometer (MIVIS) airborne data and multispectral images, such as Quickbird and Ikonos, were used. From scatter plots obtained by field data using λ = 460 and 740 nm, referring to MIVIS Bands 2 and 16 and Quickbird and Ikonos Bands 1 and 4, pixels corresponding to asphalt covering were identified, and the slope of their interpolation lines, assumed as asphalt lines, was calculated. These slopes, used as threshold values in the Spectral Angle Mapper (SAM) classifier, obtained an overall accuracy of 95% for Ikonos, 98% for Quickbird and 93% for MIVIS. Laboratory investigations confirm the existence of the asphalt line also for new asphalts, too.
Highlights
Knowledge of asphalt spectral characteristics plays an important role in several applications, such as imagery calibration and validation, land cover analysis and civil engineering
The classifications obtained for the 3 Multispectral Infrared and Visible Imaging Spectrometer (MIVIS) images present an overall accuracy (OA) between
Field spectra are used to compute the equation of the asphalt line for each area, and the angular coefficients (α) show values from 0.6 to 0.8
Summary
Knowledge of asphalt spectral characteristics plays an important role in several applications, such as imagery calibration and validation, land cover analysis and civil engineering. The application use of spectrally non-variant targets, such as asphalts, facilitates the atmospheric correction of satellite imagery [1,2,3] These pseudo-invariant targets can be used for image calibration and validation, but uncertainties about their classification may yield erroneous results. The ability to classify these surfaces, according to their chemical-physical components, alterations and level of bitumen coverage, may be essential for endmember selection Due to their geometrical and spectral characteristics, paved surfaces are recognizable in images and often represent fundamental targets for multi-temporal analysis [4] and urban land cover studies [5]. Asphalt classification by remote sensing could be a useful approach to optimize road network management. Recent studies [6,7,8,9] show a growing interest in remote and proximal sensing applications for road network monitoring and management
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
