You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Fusion of sensor data, knowledge sources and algorithms for extraction and classification of topographic objects
Baltsavias, Emmanuel P.

International Archives of Photogrammetry and Remote Sensing, Vol. 32, Part 7-4-3 W6, Valladolid, Spain, 3-4 June, 1999
2000 .
500 -,
Fig. 3. Integration result of both smooth DTMs. The integrated estimates at multiple scales are plotted.
low resolution level. This is because the accuracy of the second
DTM is higher than the accuracy of the first one. The computed
reconstruction using the information of the two DTMs in each
experiment can be seen in Figures 3 and 5, respectively.
Already visually a reduction of the noise in the integrated
DTMs can be observed. This means, that, independently of the
terrain type, the second, lower resolution DTM with the more
precise height values in its raster points contributes through the
wavelet reconstruction in such a way that the noisy high
resolution DTM after integration receives a significant
improvement on all scale levels. For the wavelet transform,
Daubechies wavelets of order 2 (Daubechies, 1992) are used,
even though the use of the simpler Haar wavelet (which in
general lacks smoothness) would not result in a visually
recognisable difference.
Fig. 4. Experiment with undulating terrain. Plotted is the high resolution DTM with three coarser scale levels.