Share Email Print
cover

Proceedings Paper

Aerial photography restoration using the maximum likelihood estimator (MLE) algorithm
Author(s): Ignacio Juvells Prades; Jorge Nunez; Fernando Perez; Vincenc Pala; Roman Arbiol
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

We apply the maximum likelihood estimator (MLE) method to restore scanned photogrammetric plates. In color images, the restoration is carried out separately for each band. To compute the MLE solution, we use the algorithm based on the expectation maximization (EM) algorithm for Poisson data. The memory and CPU time needed to obtain MLE solutions make feasible the restoration of images of 512 X 512 or 1024 X 1024 pixels, but the restoration of a series of large scanned photographs, as happens with photogrammetric plates, is not of practical use with present-day computers. With the aim of bypassing this problem, several 512 X 512 portions of the plate are extracted and restored with the MLE algorithm. Then, the convolution function transforming the original into the restored image is calculated in Fourier space, thus obtaining a convolution matrix when returning to normal space. Next, this matrix is truncated and a convolution is performed over the whole image. A series of tests over the same image digitized with different scanners has been carried out to separate this contribution from environmental effects.

Paper Details

Date Published: 17 August 1994
PDF: 6 pages
Proc. SPIE 2357, ISPRS Commission III Symposium: Spatial Information from Digital Photogrammetry and Computer Vision, (17 August 1994); doi: 10.1117/12.182866
Show Author Affiliations
Ignacio Juvells Prades, Univ. de Barcelona (Spain)
Jorge Nunez, Univ. de Barcelona and Observatorio Fabra (Spain)
Fernando Perez, Institut Cartografic de Catalunya (Spain)
Vincenc Pala, Institut Cartografic de Catalunya (Spain)
Roman Arbiol, Institute Cartografic de Catalunya (Spain)


Published in SPIE Proceedings Vol. 2357:
ISPRS Commission III Symposium: Spatial Information from Digital Photogrammetry and Computer Vision
Heinrich Ebner; Christian Heipke; Konrad Eder, Editor(s)

© SPIE. Terms of Use
Back to Top