context of the scale-space theory for signal processing. The original method had been
ranked among the top three most accurate methods of a number of previously
developed methods (Silván-Cárdenas and Wang, 2006). In this work, we further
improved the method by incorporating several novel features, including (1) the
generalization of the erosion operator for shifting not only the zero-order coefficient
but also the first-order coefficients in a rotated framework, (2) relaxing the assumption
on maximally flat terrain underneath nonground features by allowing linear prediction
of coefficients at these points through a truncated Taylor expansion, (3) the replacement of the symmetric condition to an antisymmetric condition when computing the
MDHT coefficients along boundary cells, and (4) a mechanism to refine the DTM
through repeated application of the multiscale erosion operator. In addition, practical
considerations on parameter selection were provided.
The new method was compared with the original method and results showed
improvement along areas where terrain is particularly complex or sloppy. The method
had been shown to be robust on low-resolution data sets and able to effectively
remove large buildings of complex size. However, further testing is necessary to
automatize the parameter selection and to determine in advance an appropriate
number of interactions. Also, it has been claimed that the raster-based approach
adopted here is more efficient than the point-based approach; however, a comparative
study on time complexity of ground filtering algorithms is still lacking.
REFERENCES
Axelsson, P. 1999. Processing of laser scanner data—algorithms and applications. ISPRS
Journal of Photogrammetry & Remote Sensing 54:138–147.
Axelsson, P. 2000. DEM generation from laser scanner data using adaptive TIN models.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information
Sciences 33:110–117.
Brovelli, M. A., Cannata, M., and Longoni, U. M. 2002. Managing and processing
LIDAR data within GRASS. GRASS Users Conference. University of Trento, Trento,
Italy: 29 pp.
Chen, Q., Gong, P., Baldocchi, D., and Xie, G. 2007. Filtering airborne laser scanning data with
morphological methods. Photogrammetric Engineering & Remote Sensing 73(2):175–185.
Evans, J. S., and Hudak, A. T. 2007. A multiscale curvature algorithm for classifying discrete
return lidar in forested environments. IEEE Transactions on Geoscience and Remote
Sensing 45(4):1029–1038.
Felsberg, M., and Sommer, G. 2004. The monogenic scale-space: A unifying approach to
phase-based image processing in scale-space. Journal of Mathematical Imaging and Vision
21(1):5–26.
Habib, A. 2009. Accuracy, quality assurance, and quality control of LiDAR data. In J. Shan and
C. K. Toth (Eds.) Topographic Laser Ranging and Scanning, Boca Raton, FL: CRC Press,
pp. 269–294.
Iijima, T. 1959. Basic theory of pattern observation. Papers of Technical Group on Automata
and Automatic Control. Japan: IECE.
REFERENCES
283
ranked among the top three most accurate methods of a number of previously
developed methods (Silván-Cárdenas and Wang, 2006). In this work, we further
improved the method by incorporating several novel features, including (1) the
generalization of the erosion operator for shifting not only the zero-order coefficient
but also the first-order coefficients in a rotated framework, (2) relaxing the assumption
on maximally flat terrain underneath nonground features by allowing linear prediction
of coefficients at these points through a truncated Taylor expansion, (3) the replacement of the symmetric condition to an antisymmetric condition when computing the
MDHT coefficients along boundary cells, and (4) a mechanism to refine the DTM
through repeated application of the multiscale erosion operator. In addition, practical
considerations on parameter selection were provided.
The new method was compared with the original method and results showed
improvement along areas where terrain is particularly complex or sloppy. The method
had been shown to be robust on low-resolution data sets and able to effectively
remove large buildings of complex size. However, further testing is necessary to
automatize the parameter selection and to determine in advance an appropriate
number of interactions. Also, it has been claimed that the raster-based approach
adopted here is more efficient than the point-based approach; however, a comparative
study on time complexity of ground filtering algorithms is still lacking.
REFERENCES
Axelsson, P. 1999. Processing of laser scanner data—algorithms and applications. ISPRS
Journal of Photogrammetry & Remote Sensing 54:138–147.
Axelsson, P. 2000. DEM generation from laser scanner data using adaptive TIN models.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information
Sciences 33:110–117.
Brovelli, M. A., Cannata, M., and Longoni, U. M. 2002. Managing and processing
LIDAR data within GRASS. GRASS Users Conference. University of Trento, Trento,
Italy: 29 pp.
Chen, Q., Gong, P., Baldocchi, D., and Xie, G. 2007. Filtering airborne laser scanning data with
morphological methods. Photogrammetric Engineering & Remote Sensing 73(2):175–185.
Evans, J. S., and Hudak, A. T. 2007. A multiscale curvature algorithm for classifying discrete
return lidar in forested environments. IEEE Transactions on Geoscience and Remote
Sensing 45(4):1029–1038.
Felsberg, M., and Sommer, G. 2004. The monogenic scale-space: A unifying approach to
phase-based image processing in scale-space. Journal of Mathematical Imaging and Vision
21(1):5–26.
Habib, A. 2009. Accuracy, quality assurance, and quality control of LiDAR data. In J. Shan and
C. K. Toth (Eds.) Topographic Laser Ranging and Scanning, Boca Raton, FL: CRC Press,
pp. 269–294.
Iijima, T. 1959. Basic theory of pattern observation. Papers of Technical Group on Automata
and Automatic Control. Japan: IECE.
REFERENCES
283