Proceedings of the Symposium on Global and Environmental Monitoring: Proceedings of the Symposium on Global and Environmental Monitoring

Bibliographic data

Multivolume work

Persistent identifier:: 856665355

Title:: Proceedings of the Symposium on Global and Environmental Monitoring

Sub title:: techniques and impacts ; September 17 - 21, 1990, Victoria Conference Centre, Victoria, British Columbia, Canada

Year of publication:: 1990

Place of publication:: Victoria, BC

Publisher of the original:: [Verlag nicht ermittelbar]

Identifier (digital):: 856665355

Language:: English

Document type:: Multivolume work

Volume

Persistent identifier:: 856669164

Title:: Proceedings of the Symposium on Global and Environmental Monitoring

Sub title:: techniques and impacts; September 17 - 21, 1990, Victoria Conference Centre, Victoria, British Columbia, Canada

Scope:: XIV, 912 Seiten

Year of publication:: 1990

Place of publication:: Victoria, BC

Publisher of the original:: [Verlag nicht ermittelbar]

Identifier (digital):: 856669164

Illustration:: Illustrationen, Diagramme, Karten

Signature of the source:: ZS 312(28,7,1)

Language:: English

Usage licence:: Attribution 4.0 International (CC BY 4.0)

Editor:: International Society for Photogrammetry and Remote Sensing, Commission of Photographic and Remote Sensing Data

Publisher of the digital copy:: Technische Informationsbibliothek Hannover

Place of publication of the digital copy:: Hannover

Year of publication of the original:: 2016

Document type:: Volume

Collection:: Earth sciences

Chapter

Title:: [WP-4 INTERPRETATION AND MODELLING]

Document type:: Multivolume work

Structure type:: Chapter

Chapter

Title:: AUTOMATED SYSTEM FOR IMAGE INTEGRATION IN REMOTE SENSING. Malcolm E. Fuller & Manfred Ehlers

Document type:: Multivolume work

Structure type:: Chapter

O) CD <D J* O <D Cl Reference image Subject image Figure 2. "La Grange" flow chart The program EVEN MORAVEC performs the CCP selection in the images using the Moravec operator as modified. The input for the program consists of the image file to extract CCPs from. The program must be run separately for the reference and subject images. Output is a text file containing a list of the candidate point image coordinates (line and pixel) and their interest values. The program TIN HORN is a correspondence program and operates in the "quick and dirty" fashion mentioned above. An improved correspondence strategy is currently being investigated which will avoid the foibles of the current algorithm. TIN HORN may, however, have a place in future versions of "La Grange," either in a modified version or as a complementary component of a more sophisticated correspondence scheme. The input for TIN HORN are the two CCP lists, one from the reference image and the other from the subject image, which have been created by EVEN MORAVEC. The output is a single text file containing the CCPs in each image which have been matched. HARDCORR is a program which correlates the CCPs to fine tune their locations in the subject image. It is derived from an already existing correlation package whose purpose was to generate parallax files from stereo images [Ehlers and Welch 1987]. It takes as input the reference and subject images and the matched CCP list created by TIN HORN. The output is an ERDAS-readable GCP file containing the correlated control points. This file can be passed to the ERDAS image processing module for calculation of the transformation polynomial coefficients and resampling of the subject image. In addition to those programs shown in figure 2, there is also a display program to allow the user to visually inspect the locations and appearance of the CCPs after the selection, matching, and fine tuning steps have been performed. This is a useful tool for testing purposes and to demonstrate visually the process of CCP refinement to a final set of control points. PRELIMINARY RESULTS Each of the "La Grange" component programs have been tested individually. The source code from which EVEN MORAVEC and HARDCORR were written are known to have been thoroughly tested in their original form. Therefore, these programs have been tested mainly for debugging purposes. Program TIN HORN performs as well as expected (i.e. marginally) in tests. Combined testing of the entire "La Grange" package has not been thoroughly performed as of this report. Some initial testing has been done, however, using a small (256x240) subset of a Landsat TM band 5 image of an area near Tiger, GA. This dataset was used as the reference image and a slightly skewed (rotated clockwise approximately 8 pixels at each corner) copy of the it was made for use as a subject image (keeping in mind the limitations of TIN HORN). "La Grange" was then used to attempt to restore the skewed image to its original form. The following shows the results at each step of the matching process given the default user- input parameters:

Access restriction

Copyright

fullscreen: Proceedings of the Symposium on Global and Environmental Monitoring (Part 1)

Multivolume work

Volume

Chapter

Chapter

Table of contents

Cite and reuse

Volume

Chapter

Image

Image fragment

Citation links

Citation recommendation