Photogrammetric Mapping

A case study of El Salvador City has been done using various photogrammetric techniques. This project comprised the complete new mapping of the town including all features like buildings, boulevards, roads,manhole/ drainage, light / telephone polls, trees /tree areas, ponds, river and streams (with break lines), Monuments, playground / Parks etc. The vectorization of the model has been done on Digital (Socket Set) and analytical systems both. From speed point of views digital systems are as good as analytical systems. It also depends on the skills / expertise of the Operator who is preparing the maps.

Material used Two Diapositive with a scale of 1:5000, Two scanned images with 14 micron per pixel and two images with a resolution of 24 micron per pixel. Scale of Photographs: 1: 5000

Methodology

An appropriate method for assessing the accuracy of derived data and the impact of parameters on the data is to carry out tests using independent check data in the object space. Quality control, or checkpoints with a known elevation considered to be true, were used for accuracy assessments and were measured using DGPS (differential global positioning system). Diapositives are used in Analogue (A8 & AMH) and Analytical plotters (BC3, SD2000).

To start a project on analogue/ analytical systems, Ground Control Points (CP File) is required or created for the entire area / model. In these plotters one MD ?le (orientation parameter ?le) is created for every model using control points.

In Analogue Machines, orientation is totally manual. Inner Orientation is done on a light table where four ?ducial points are brought in ?ducial axis marked on the picture carrier plate on which the diapositive is mounted. Later, it is mounted on instrument for RO and AO. In Relative
Orientation Y-parallax are removed using Kappa, Phi and Omega. After that Absolute orientation is performed where Known GCP’s are measured. It takes more than half an hour to orient a single model.

In Analytical Plotters, orientation is automatic. For making orientation parameter ?les (MD Files), ?rst of all, diapositives were mounted on instrument and started with Inner Orientation. It starts with left image followed by right image. These are automatic instrument even then coordinates at ?rst two ?ducial points were measured manually. Later it reads remaining six ?ducial points automatically. The same process is done for right photograph. The RMS value of IO should be less than 10. Epipolar Orientation follows inner Orientation where parallax is removed manually at two points. After the Epipolar orientation, Relative Orientation is performed. During RO process, Y parallax is removed at least at six points or maximum at 10 points. For Absolute Orientation, ground control points are measured and RMS is noted. After AO process, model becomes ready for data extraction / map production.

For Digital Photogrammetry, photographs were scanned, using an LH System’s photogrammetric scanner, at a resolution of 14 microns and 24 microns per pixel prior to reformatting and transfer into the Socket set Digital Photogrammetric Workstation (DPW). Full details of camera calibration parameters were supplied to enable interior orientation to be carried out. All processes like IO, RO and AO are done for digital systems also. Stereo model were constructed using premarked ground control points positioned using post-processed differential GPS. Subsequent analysis involved the calculation of residual values produced by comparing the computer-generated surface with a set of test points measured using differential GPS. This research demonstrates that the optimal digital system matched the performance of analytical photogrammetry for the feature extraction.

Observations

After AO process, RMS is checked. About 20 points were observed in Analytical and Digital system. Later these points were compared with known GCP’s.

Following RMS values are obtained in Analytical system:

Planimetry = 14.0 cm

Altimetry = 10.0 cm

Following RMS values are obtained in Digital system:

Planimetry = 17.0 cm

Altimetry = 12.8 cm

The above result is from photographs with a scale of 1:5000

Note: – If photographic scale is small like 1:25000, then RMS may be different. Planimetry accuracy may go around 50 cm and Altimetry accuracy may go around 80 cms.