Traditional specifications
for aerial LIDAR call for one pulse per square meter point density to derive the ubiquitous one foot contours as specified
by the FEMA specs for floodplain mapping, however, a host of other applications require much higher point densities.
Increasingly, LIDAR customers are requesting higher point densities for a
variety of reasons. They have found that by collecting higher point densities by flying lower and slower they are able to
saturate the ground with 5 to 20 even up to 40 points per square meter. That means an object can be hit by multiple pulses
and increases the chances of hitting the ground in areas of dense vegetation.
Flying at around 400 meters above mean terrain at 120 Knots collecting 200,000 measurements per second, insures greater
ability to measure the ground through heavy vegetation and provides a more accurate and better overall map product. This method
is increasingly being used for topographic surveys of areas with vegetation, during leaf-on conditions to produce one-foot
contours.
The holy grail of object identification is driving the need for higher point
densities, where the level and precision of object identification depends greatly on the density of the laser point
cloud.
Measurement overlap and optimal ground coverage provide a number of distinct advantages such as the ability to detect
small linear objects and dramatic changes in elevation caused by planimetric features such as retaining walls, ditches, ridges
and embankments. Proper data filtering provides the ability to effectively identify buildings and roof lines, or drainage
patterns and other hydrographic details. This level of detail enables diverse application potential, such as urban planning
and the production of 3D city modeling where cadastral and engineering data can be integrated with rendered building structures
to visualize expansion plans, redevelopment projects and infrastructure networks.
High-density LIDAR data is proving valuable to forest management. In addition to providing a much higher level of
accuracy of the bare-earth, it also enables single tree segmentation, crown area verification, trunk diameter measurements
and tree center coordinate values. In areas where open foliage is limited, wide beam, narrow viewing angle, and optimal ground
coverage vastly increase the probability of canopy penetration.
The trend toward higher point densities
has been driven extensively by the electric transmission market where the demand is for detailed and precise vectorized data
of power lines, conductors and structures. Here the objective is the extraction and vectorization of individual power lines
as well as 3D coordinates such as pylon positions, cable fixations, attachment points, polygon center lines, lower cross arm
edges, power line sag and several other details.
Hardware improvements
have come a long way in accelerating the use of such high-density LIDAR models. Fortunately, software has
also improved. As the demand for technology advancements to increase point densities has increased, software
advances have had to adapt to the ever increasing amounts of data. The more advanced LiDAR software offerings
incorporate architectures that handle entire point clouds without memory limitations and they leverage the industry standard
LAS format to provide access to all available LiDAR attributes.
Accessing the LAS file
directly is important as it provides opportunities to utilize new and advancing classification and feature extraction techniques.
Such advancements are increasingly relying on information contained in the LAS file. Unfortunately,
previous generations of software often used proprietary imported file solutions that stripped the point cloud information
in order to obtain a smaller file size, thus limiting the ability to apply new filtering and extraction techniques to these
datasets.
The LiDAR users of today want to be able to leverage high-density LAS point clouds in an efficient and effective
manner. An example of a software package pushing higher point density requirements is QCoherent Software’s LP360. Using a non-RAM sensitive approach to accessing LiDAR, LP360 allows the rich, dense, LAS LiDAR point
clouds of today to be seamlessly accessed in ArcGIS. Functional data analysis, feature identification,
conflations, and classification of LiDAR with LP360 are all made possible in ArcGIS.
Software solutions that combine legacy data with the high point densities of today’s LiDAR collections are
fueling the accelerated use of LiDAR in geospatial analysis. This in turn is further driving the demand
for higher point densities. Much like demand for higher resolution imagery, the push for higher density point clouds will
continue into the foreseeable future.
By the LIDARcomm staff
For additional information ...email info@lidarcomm.com.
