The most modern development in aerial photography is the spurt of using unmanned aerial vehicles for image capturing. This has reduced the image acquiring cost to a great extent that too in extremely small time spans. The added advantages of UAV image capturing are its ability to take very high resolution images and it’s on board storage. It has revolutionized the application of UAV data usage in disaster management, precision farming, and topographical data collection of mining areas, solar farms, and border security, fishing surveillance, traffic management and many more. Though it has many advantages, this type of photography has brought its own challenges in the processing of the UAV acquired images. The major challenges posed by this technology in data processing are briefed below.
Inconsistent over lapping: This is a major challenge in UAV data processing. This will leads to non generation of stereo models and if generated collapsing of features. The author’s technical experience shows that a minimum of 80% overlap is required along flight line and 70% overlap is required between flight lines.
Distribution of Ground Control Points: Accuracy and distribution of ground control points are another challenge in UAV data processing. When comparing with conventional aerial photography, UAV require more number of Ground Control Points (GCP). As UAV data collection will be generally for small areas a dense distribution of GCPs are required for high accuracy. These GCPs should be evenly distributed and cover all the borders adequately. Accuracy of the GCPs is a defining factor in accuracy of digital terrain models and derived products.
Requirement of hardware and software: UAV data processing requires additional software to process the data. There are quite few software available in the market for processing UAV data. Each software has its own merit and demerits. Most of the software requires high end hardware to work on. This is a major challenge in data processing. As UAV swath is very small, number of images will be very high. In such instances one should require a high end hardware system with high level of RAM and very good quality graphic card, high storage space and other peripherals.
Block to Block changes in altitudes: As UAVs are very small in size and flying at a lower altitude comparing with conventional aerial photography, causes flight line changes in altitude. This will leads to more manual editing along the block boundaries.
Failure areas: UAV data contain many numbers of images even for a small area. So, almost all software output will show failure areas much greater in amount than conventional photogrammetric methods. So a sizable amount of manual interference is required for correcting the Digital Surface Models (DSM). Data processing involves tedious manual editing of the failed and collapsed areas.
Absence of Auto Digital Terrain Models (DTM): None of the existing well established software are not providing with direct DTM. So manual editing with break lines is essential for getting very high accuracy Digital Terrain Models (DTM).
Issues in ortho photos: Another challenge in UAV data processing is the collapsed areas in ortho mosaics. Some of the software claims that they have this auto correction facility, but in practical manual interference is required. Ortho photo mosaic editing is required using Photoshop which is a manual tedious process.
SBL Geomatics provides complete solutions for all standard tasks in digital photogrammetry services such as aerial triangulation, DEM / DTM generation, photogrametry services, stereo compilation, topographic and planimetric feature extraction, orthophoto production, color balancing, mosaicking and tile generation, pan sharpening, contour generation, 3D terrain visualization, orthophotography services and LiDAR services
No comments:
Post a Comment