Lab 3: Vegetation Classification

Goals and Background

The purpose of the lab is to gain experience classifying Lidar data into various vegetation height classes. Accurate classification of vegetation can allow the analyst the ability to extract vegetation metrics which are used for numerous applications. I will be utilizing height filters to apply an algorithm to classify the majority of the remaining unclassified points and then finishing the take using manual clean up methods.

Methods

 All of the following methods were performed in LP360. I will be utilizing the data from labs 1 & 2 where I classified the ground and building points.

Vegetation Classification 

I created a new Macro task in the Point Cloud Tasks tab (Fig 1). The classification was labeled Vegetation Classification. The task type was set to Height Filter and labeled Vegetation Extraction. The source points were set to Unclassified and the ground points were set to Ground and Reserved (Model keypoints). Next I added the height filter ranges per the instruction of my professor. The ranges were as follows:

• >1.640-<= 6.6 (Low Vegetation)
• >6.6-<=19.7 (Medium Vegetation)
• >19.7-<=328 (High Vegetation)
• >328-Open (High Noise)

All flags were ignored for this classification.

(Fig. 1) Vegetation extraction parameter settings in the Point Cloud Tasks window.

Manual Cleanup

Utilizing the profile window I examined the vegetation classification (Fig. 2). In the example shown in figure 2 you can see the classification was performed fairly well. You will notice a few red (building classification) points in a few areas. I utilized the Small Paint Brush tool to select these points. Utilizing the selecting filter I set my source points to Building. Then I changed the Destination Class & Flags to the appropriate vegetation class based on the height requirements from the above classification parameters. After selecting the Building classified points I used the Space Bar to run the filter to change them their proper vegetation class (Fig. 3).

(Fig. 2) Vegetation classification of a sample area using the profile tool before correcting the results.

(Fig. 3) Vegetation classification of a sample area using the profile tool after performing the correction

Results

(Fig. 4) Display of a sample area containing results from my classification.

Examining figure 4 you can see numerous errors in the classification which will have to be fixed to have a proper classification of the study area. Lab 4 will have me performing QA/QC on the data and utilizing filters to correct many of the error which have been creating in the original classification.

Sources

LAS, tile index, and metadata for Lake County are from Illinois Geospatial Data ClearingHouse. NAIP imagery is from United States Department of Agriculture Geospatial Data Gateway. Breaklines is from Chicago Metropolitan Agency for Planning.