The impervious map was created using “expert systems” rulesets developed in Trimble Ecognition. These rulesets combine automated image segmentation with object based image classification techniques. In contrast with machine learning approaches, expert systems rulesets are developed heuristically based on the knowledge of experienced image analysts.  Key data sets used in the expert systems rulesets for impervious mapping included:  orthophotography (’11 and ’13), the LiDAR point cloud, and LiDAR derived rasters.

After it was produced using Ecognition, the preliminary impervious map product was manually edited by a team of photointerpreters. Manual editing corrected errors where the automated methods produced incorrect results.

The impervious surfaces map reflects the state of the landscape in fall, 2013 when countywide high resolution imagery and LiDAR collection occurred.

The impervious map has 4 classes, which are described below:

• Buildings - All buildings above 200 square feet in area.

• Dirt or Gravel Roads - Dirt or gravel roads wide enough for vehicle traffic. Dirt trails too narrow for a vehicle should be considered pervious.

• Paved Roads - Roads that are paved, regardless of pavement condition.

• Other Impervious - Impervious areas that are not paved roads, dirt roads, or buildings. Includes parking lots, sidewalks, paved walking paths, swimming pools, tennis courts, and man-made areas of compacted soil (such as dirt parking areas, pull offs along roads, and compacted areas near commercial activities). This class DOES NOT include baseball diamonds, very lightly used dirt roads, foot trails, livestock areas, naturally occurring bare soil or rock, or bare areas around ponds.

Miscellaneous quality control and processing notes:

• Zoom level used during manual quality control was no finer than 1 to 2,000.

• Vector data was created with no overlapping polygons.

Minimum Mapping Units:

The table below shows the nominal minimum mapping units (MMUs) for the impervious surfaces map.

The impervious map was created using “expert systems” rulesets developed in Trimble Ecognition. These rulesets combine automated image segmentation with object based image classification techniques. In contrast with machine learning approaches, expert systems rulesets are developed heuristically based on the knowledge of experienced image analysts.  Key data sets used in the expert systems rulesets for impervious mapping included:  orthophotography (’11 and ’13), the LiDAR point cloud, and LiDAR derived rasters.

After it was produced using Ecognition, the preliminary impervious map product was manually edited by a team of photointerpreters. Manual editing corrected errors where the automated methods produced incorrect results.

The impervious surfaces map reflects the state of the landscape in fall, 2013 when countywide high resolution imagery and LiDAR collection occurred.

The impervious map has 4 classes, which are described below:

• Buildings - All buildings above 200 square feet in area.

• Dirt or Gravel Roads - Dirt or gravel roads wide enough for vehicle traffic. Dirt trails too narrow for a vehicle should be considered pervious.

• Paved Roads - Roads that are paved, regardless of pavement condition.

• Other Impervious - Impervious areas that are not paved roads, dirt roads, or buildings. Includes parking lots, sidewalks, paved walking paths, swimming pools, tennis courts, and man-made areas of compacted soil (such as dirt parking areas, pull offs along roads, and compacted areas near commercial activities). This class DOES NOT include baseball diamonds, very lightly used dirt roads, foot trails, livestock areas, naturally occurring bare soil or rock, or bare areas around ponds.

Miscellaneous quality control and processing notes:

• Zoom level used during manual quality control was no finer than 1 to 2,000.

• Vector data was created with no overlapping polygons.

Minimum Mapping Units:

The table below shows the nominal minimum mapping units (MMUs) for the impervious surfaces map.