Peer Reviewed Articles
687-694 Tone Reproduction of Photographic Scanners
O. Kölbl and U. Bach
Abstract
Download
Full Article
The aims of this study were to analyze the tone reproduction of photographic
scanners and to develop simple test procedures. The most important arguments
proposed are the image noise, image resolution, sensitivity of the scanner,
and the visual aspect of the images. A good scanner should show an image noise
lower than ± 0.03 to 0.05 D for a pixel size of 10µm and a resolution
up to 10µm. Furthermore, the dynamic density range and the fidelity
of tone reproduction should be controlled. The study includes tests of the
scanners most commonly used in photogrammetry and in the printing industry,
but shows that their efficiency does not allow them to fully exploit the
high image quality of modern aerial photographs.
695-701 A Comparison of Two Image Compression Techniques for Softcopy Photogrammetry
Kurt Novak and Fayez S. Shahin
Abstract
Download
Full Article
This paper discusses the basics of digital image compression and explains
the conceptual differences of various approaches. Two techniques, the
standard JPEG algorithm and the newly developed Hierarchical Predictive Coding
(HPC) approach, are described in detail. They are compared with respect
to geometric quality, loss of information, compression ratio, and compression
speed. JPEG yields better compression ratios; however, it is significantly
slower than xPC.
703-710 Automated Aerotriangulation Using Multiple Image Multipoint Matching
Peggy Agouris and Toni Schenk
Abstract
Download
Full Article
This paper deals with automating the aerotriangulation procedure using digital
image processing methods and images available exclusively in digital format.
A conceptual system has been developed and implemented which allows for the
automatic determination of exterior orientation parameters and positions
of selected points in object space using a minimal number of assumptions.
In direct analogy to the preparation phase of the conventional approach,
a photomosaic is automatically generated to describe the interrelationship
of all photos within the available block. This photomosaic is subsequently
used for the automatic extraction of approximate conjugate image coordinates.
A multiple image multipoint matching strategy has been developed to identify and precisely match conjugate points in the triple, quadruple, or multiple overlapping areas of the block. By taking into account the geometric relationship between conjugate points as expressed by their corresponding image forming rays, and using image coordinates rather than the widely used local template coordinates, matching and block adjustment can be combined in a unique way. Thus, the mensuration and adjustment phases of classical aerotriangulation are substituted by a single digital process which allows automation of the complete procedure.
711-717 Concept, Implementation, and Results of an Automatic Aerotriangulation
System
Charles K. Toth and Amnon Krupnik
Abstract
Download
Full Article
During the last two decades, aerotriangulation has enjoyed great success.
Block adjustment programs and analytical plotters have contributed the
most to reliable and accurate triangulation procedures. Digital photogrammetry
and softcopy workstations have added a new dimension to aerotriangulation,
and some of the previously manual procedures can now be automated. In
this paper, we describe a softcopy-based automatic aerotriangulation system
that generates the block configuration, selects block points (tie points),
and matches them to sub-pixel accuracy. A key element is the sD block system
which contains information about the coverage of photographs.
719- A Digital System for Surface Reconstruction
Weiyang Zhou, Robert H. Brock, and Paul F. Hopkins
Abstract
Download
Full Article
A digital photogrammetric system, STEREO, was developed to determine three-dimensional
coordinates of points of interest ; (POIs) defined with a grid on a textureless
and smooth-surfaced specimen. Two CCD cameras were set up with unknown
orientation and recorded digital images of a reference model and a specimen.
Points on the model were selected as control or check points for calibrating
or assessing the system. A new algorithm for line feature detection called
local maximum convolution (LMC) helped extract the POIs from the stereo image
pairs. The system then matched the extracted POIs and used a least-squares "bundle" adjustment
procedure to solve for the camera orientation parameters and the coordinates
of the POIs. An experiment with STEREO found that the standard deviation
of the residuals at the check points was approximately 24 percent, 49 percent,
and 56 percent of the pixel size in the X, Y, and Z directions, respectively.
The average of the absolute values of the residuals at the check points
was approximately 19 percent, 36 percent, and 49 percent of the pixel size
in the X, Y, and Z directions, respectively. With the graphical user interface,
STEREO demonstrated a high degree of automation, and its operation does not
require special knowledge of photogrammetry, computers, or image processing.
727 - Digital Terrain Model
(DTM) from Indian Remote Sensing (IRS) Satellite Data from the Overlap
Area of Two Adjacent Paths Using Digital Photogrammetirc Techniques
T. Ch. Malleswara Rao, K. Venugopala Rao, A. Ravi Kumar, D. P. Rao, and
B. L. Dekkshatula
Abstract
Download
Full Article
Present day needs for medium- and small-scale topographic mapping at
reduced cost and time can be met by using stereoscopic digital data
acquired by present and planned satellites. In anticipation of the planned
IRS-1C
stereo data, attempts are made to use the overlap area between two adjacent
paths of IRS-1A linear imaging self scanning (LISS)-II data as a stereo
pair for generating a digital terrain model (DTM) using digital photogrammetric
techniques. Software has been developed to recover the epipolar geometry
of the stereo pair and to accomplish automated digital stereo matching
to find the corre¬sponding points in the left and right images by correlation
techniques to calculate parallax to the sub-pixel level. The parallax is
then used in conjunction with the IRS-1A data acquisition system to
compute relative elevations. The absolute DTM is generated by referring
the relative elevation to a few control points with known Z-coordinates.
To evaluate
the de¬veloped methodology and the software, three test sites at different
locations in India near Palghat, Pune, and Dehradun were selected. It is
observed that the overlap area between two adjacent paths increases from
the equator to the poles whereas the base-to-height ratio (B/H) decreases.
A methodology to calculate the percentage of overlap and B/H value at a
given latitude is presented. The resultant DTMS for the three test sites
were evaluated
for error, and it is observed that 90 percent of RMSz errors fall within
33.95 m, 34.45 m, and 38.12 m at Palghat, Pune, and Dehradun, respectively.
The DTM and the corresponding IRS-1A data are used to generate sD perspective
images. The software package developed will have potential usage in sD
analysis of IRS-1C stereo data
733 The Digital Transferscope
Eugene E. Derenyi
Abstract
Download
Full Article
A simple scheme is presented for the real-time geometric coregistration
of digital maps and images, in support of transferring
planimetric details from image to map. The essence
of this scheme is to transform the vector-based map graphics
by incremental translation, scaling, rotation, and skewing to
fit the raster-image backdrop, within a segment of the image
at a time. The newly digitized features are concatenated with
the base map by an inverse transformation. This scheme, in
fact, emulates digitally the operation of a Zoom Transferscope.
The utility of the scheme has been tested in a map revision
experiment with satisfactory results.