Below is a collection of some of the most frequently asked questions, and their answers, regarding the LIDAR Analyst® software. If you have additional questions, feel free to contact VLS via email at support@vls-inc.com.

General Questions:

Q: What formats for the input LIDAR data does LIDAR Analyst support?

Q: How accurate is LIDAR Analyst in its extraction processes?

Q: Since the evaluation version only allows for a 1000 x 1000 pixel image, how can I best attain that image size?

Technical Questions:

Q: What are the system requirements for machines running LIDAR Analyst?

Q: Does LIDAR Analyst read .LAS or point cloud information?

Q: Does LIDAR Analyst use only its raster output images and not the original points stored in the LAS file for the information extraction (bare earth, buildings, trees)?

Q: Regarding the size/resolution of LIDAR images, how many pixels per building are necessary for a good extraction? In other words, if the smallest building to extract is 9x9 meters, what should be the density of the point cloud and the resolution of the input First/Last return in LIDAR Analyst?

Q: What is the optimal resolution of LIDAR images for trees and forest extraction (relationship of tree size to point spacing)?

Q: The parameters you suggest for the building extraction are
optimal for a resolution of 1m, right? Did you also test with other resolutions (0.5 or 2 m) and do you have some recommendations for these cases?

General Questions:

Q: What formats for the input LIDAR data does LIDAR Analyst support?

A: LIDAR Analyst supports raster data, GeoTiff, IMG, LAS, and ASCII formats.

Q: How accurate is LIDAR Analyst in its extraction processes?

A: Based upon 3,100 surveyed ground points from ERDC-TEC: (1) bare earth/leaf off is equivalent to the sensor; (2) bare earth/leaf on is 6cm difference; (3) building extractions are 97% accurate after extracting 11,000 buildings in 9 minutes; (4) 100,000 tree extractions in 6 minutes.

Q: Since the evaluation version only allows for a 1000 x 1000 pixel image, how can I best attain that image size?

A: Simply use the clip tool provided within LIDAR Analyst in order to minimize image file size.

Technical Questions:

Q. What are the system requirements for machines running LIDAR Analyst?

A. The minimum system requirements for LIDAR Analyst are as follows:

• Windows 2000 or XP
• 512MB of RAM or better
• Free space on the hard drive that is at least three times the size of the image being processed
• .Net 1.1 and .Net 2.0 Framework installed
• 2 GHz or better processing speed

The list below shows optimum specifications:

• The maximum amount of RAM possible (usually 2 GB)
• Dual-core processor if the machine is to be used for other tasks while Feature Analyst processing is taking place
• ATA hard drive with RAID 0

Q: Does LIDAR Analyst read .LAS or point cloud information?

A: Yes! LIDAR Analyst 4.1 provides direct point cloud LIDAR data processing and classification. However, LIDAR Analyst 3.5 still requires data be in raster or grid format.

Q: Does LIDAR Analyst use only its raster output images and not the original points stored in the LAS file for the information extraction (bare earth, buildings, trees)?

A: That is correct. LIDAR Analyst uses changes in slope in the raster layers in order to detect buildings, forests, and trees. It is essential to create a good bare earth raster in order to get good results, so the user should always generate first and last returns from the LAS file and judge their quality, and then do the same for the bare earth layer. If the goal of the user is to classify the point cloud, then he or she can step through the LIDAR Analyst workflow, generating all the outputs (bare-earth, buildings, trees and forests) and use them to classify all of the points in the LAS file.

Q: Regarding the size/resolution of LIDAR images, how many pixels per building are necessary for a good extraction? In other words, if the smallest building to extract is 9x9 meters, what should be the density of the point cloud and the resolution of the input First/Last return in LIDAR Analyst?

A: Point spacing should be less than 1.5 meters, regardless of the building size. Larger buildings can be detected if the point spacing is outside this range, but in that case the results are going to be less than ideal. The LIDAR Analyst software's algorithms have been tested extensively on data-sets with a point spacing of less than 1.5 meters. On data-sets with point spacing greater than this value, the accuracy quickly begins to deteriorate. On data-sets that have a point spacing less than 1.5 meters the accuracy increases as the point spacing decreases. We have found that point clouds with a point spacing of 1 meter are sufficient to extract most buildings in an urban data-set. The smaller a building the user wishes to extract the lower the value of point spacing should be.

Q: What is the optimal resolution of LIDAR images for trees and forest extraction (relationship of tree size to point spacing)?

A: As with buildings, point spacing of less than 1.5 meters is preferred.

Q: The parameters you suggest for the building extraction are
optimal for a resolution of 1m, right? Did you also test with other resolutions (0.5 or 2 m) and do you have some recommendations for these cases?

A: These parameters have worked on data-sets of 0.5 and 1.0 meters. The fine tuning of the parameters is not just a function of the point spacing but also the terrain characteristics. For this reason, the user should attempt to refine the parameters settings to see if a higher accuracy in extraction can be achieved for the user's terrain.