Digital Classification and Mapping of Urban Tree Cover: City of St. Paul Metadata

Full Metadata View Attributes View Sample Get Data


Digital Classification and Mapping of Urban Tree Cover: City of St. Paul



This page last updated: 09/28/2014Metadata created using Minnesota Geographic Metadata Guidelines


Metadata Summary

Originator Remote Sensing and Geospatial Analysis Laboratory, University of Minnesota
Abstract Quickbrd multispectral imagery and LIDAR were used to classify the city of St. Paul land into land cover
Browse Graphic View a sample of the data
Time Period of Content Date 2009
Currentness Reference QuickBird satellite imagery acquired on May 28, 2009
(A small stripof about 100 meters on the far west edge of the city was missing from this image and was replaced withan image acquired on June 25, 2009.)

LIDAR imagery acquired in June 2007 was available from the U.S. Army Corps of Engineers

Access Constraints The Remote Sensing and Geospatial and Analysis Laboratory, University of Minnesota, has attempted to produce accurate maps, statistics and information of land cover and impervious surface area. However, it makes no representation or warranties, either expressed or implied, for the data accuracy, currency, suitability or reliability for any particular purpose. Although every effort has been made to ensure the accuracy of information, errors and conditions originating from the source data and processing may be present in the data supplied. Users are reminded that all geospatial maps and data are subject to errors in positional and thematic accuracy. The user accepts the data “as is” and assumes all risks associated with its use. The University of Minnesota assumes no responsibility for actual or consequential damage incurred as a result of any user's reliance on the data. The data are the intellectual property of the University of Minnesota.The Remote Sensing and Geospatial and Analysis Laboratory, University of Minnesota, has attempted to produce accurate maps, statistics and information of land cover and impervious surface area. However, it makes no representation or warranties, either expressed or implied, for the data accuracy, currency, suitability or reliability for any particular purpose. Although every effort has been made to ensure the accuracy of information, errors and conditions originating from the source data and processing may be present in the data supplied. Users are reminded that all geospatial maps and data are subject to errors in positional and thematic accuracy. The user accepts the data “as is” and assumes all risks associated with its use. The University of Minnesota assumes no responsibility for actual or consequential damage incurred as a result of any user's reliance on the data. The data are the intellectual property of the University of Minnesota.
Use Constraints This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.
Distributor Organization Remote Sensing and Geospatial Analysis Lab, Univeristy of Minnesota
Ordering Instructions see website or contact infosee website or contact info
Online Linkage Click here to download data. (See Ordering Instructions above for details.) By clicking here, you agree to the notice in "Distribution Liability" in Section 6 of this metadata.


Full metadata for Digital Classification and Mapping of Urban Tree Cover: City of St. Paul


Go to Section:
1. Identification_Information
2. Data_Quality_Information
3. Spatial_Data_Organization_Information
4. Spatial_Reference_Information
5. Entity_and_Attribute_Information
6. Distribution_Information
7. Metadata_Reference_Information


Section 1 Identification Information Top of page
Originator Remote Sensing and Geospatial Analysis Laboratory, University of Minnesota
Title Digital Classification and Mapping of Urban Tree Cover: City of St. Paul
Abstract The project objective was to generate a digital land cover classification of the City of St. Paul in GIScompatible format, with emphasis on mapping the tree cover that can be used by the City to evaluate existing tree cover and potential for additional plantings. Tree cover is defined as the leaves, branches
and stems cover the ground when viewed from above.The project objective was to generate a digital land cover classification of the City of St. Paul in GIScompatible format, with emphasis on mapping the tree cover that can be used by the City to evaluate existing tree cover and potential for additional plantings. Tree cover is defined as the leaves, branches and stems cover the ground when viewed from above.
Purpose The project objective was to generate a digital land cover classification of the City of St. Paul in GIS-compatible format, with emphasis on mapping the tree cover that can be used by the City to evaluate existing tree cover and potential for additional plantings. Tree cover is defined as the leaves, branches and stems covering the ground when viewed from above.
Time Period of Content Date 2009
Currentness Reference QuickBird satellite imagery acquired on June 25, 2009

LiDAR imagery acquired in June 2007 was available from the U.S. Army Corps of Engineers

Progress Complete
Maintenance and Update Frequency None planned
Spatial Extent of Data St. Paul, Minnesota, USA
Bounding Coordinates -93.2079057
-93.0044564
44.9920419
44.8875235
Place Keywords St. Paul, Minnesota, USA
Theme Keywords Urban Tree Cover, Impervious surface, QuickBird, remote sensing
Theme Keyword Thesaurus
Access Constraints The Remote Sensing and Geospatial and Analysis Laboratory, University of Minnesota, has attempted to produce accurate maps, statistics and information of urban tree cover, land cover, and impervious surface area. However, it makes no representation or warranties, either expressed or implied, for the data accuracy, currency, suitability or reliability for any particular purpose. Although every effort has been made to ensure the accuracy of information, errors and conditions originating from the source data and processing may be present in the data supplied. Users are reminded that all geospatial maps and data are subject to errors in positional and thematic accuracy. The user accepts the data “as is” and assumes all risks associated with its use. The University of Minnesota and project affiliates assume no responsibility for actual or consequential damage incurred as a result of any user's reliance on the data. The data are the intellectual property of the University of Minnesota.The Remote Sensing and Geospatial and Analysis Laboratory, University of Minnesota, has attempted to produce accurate maps, statistics and information of urban tree cover, land cover, and impervious surface area. However, it makes no representation or warranties, either expressed or implied, for the data accuracy, currency, suitability or reliability for any particular purpose. Although every effort has been made to ensure the accuracy of information, errors and conditions originating from the source data and processing may be present in the data supplied. Users are reminded that all geospatial maps and data are subject to errors in positional and thematic accuracy. The user accepts the data “as is” and assumes all risks associated with its use. The University of Minnesota and project affiliates assume no responsibility for actual or consequential damage incurred as a result of any user's reliance on the data. The data are the intellectual property of the University of Minnesota.
Use Constraints This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.
Contact Person Information Marvin Bauer, Professor
Remote Sensing and Geospatial Analysis Lab, Univeristy of Minnesota
1530 Cleveland Avenue North
St. Paul , MN 55108
Phone: (612)624-3703
Fax: (612)625-5212
Email : mbauer@umn.edu
Browse Graphic View a sample of the data
Browse Graphic File Description
Associated Data Sets Minneapolis and Woodbury datasets also availableMinneapolis and Woodbury datasets also available


Section 2 Data Quality Information Top of full metadata Top of page
Attribute Accuracy Accuracies for the state are reported in the supplemental map file titled: St. Paul Tree Canopy Mapping - Final Report.pdf
Logical Consistency
Completeness Data provides complete coverage of St. Paul, Minnesota, USA.Data provides complete coverage of St. Paul, Minnesota, USA.
Horizontal Positional Accuracy LIDAR: The horizontal accuracy of the data was roughly 0.5 meters and stated to be “better than 1 meter.”
Vertical Positional Accuracy LIDAR: The vertical accuracy compared to 33 control points was 0.087 meters.
Lineage
The land cover classes include: tree canopy, grass and shrubs (including agricultural fields), buildings,
impervious (streets, driveways and parking areas), water and bare soil.

The primary land classifications were produced using object based image analysis (OBIA) techniques
available in eCognition Developer version 8.0. Ancillary software utilized in the project included ArcGIS
version 9.3.1 and ERDAS Imagine version 2010. Additional customized routines were written in Python
version 2.5 scripting language to support processing as required. Shapefile information was provided by
the City of St. Paul to help identify streets, buildings, roads and highways and water features.

The following principle steps were followed to implement the project:

  • QuickBird Imagery was pan sharpened using subtractive resolution in ERDAS Imagine.
  • QuickBird Imagery was georeferenced utilizing the available RPC files and a 30 meter DEM layer.
  • Lidar data was georeferenced to match the QuickBird imagery.
  • A customized Python script was used to divide the georeferenced imagery into 750 x 1000 meter tiles with 10% overlap for further processing. This step created 180 individual tiles.
  • The street layer containing road information was buffered in ArcGIS by one meter to create a polygon shapefile for subsequent use in eCognition.
  • Three rule sets were developed to process the following subsections of the city:
    • The small western section which included both June QuickBird and lidar data.
    • The 1,500 meter strip on the east side of the city which had May QuickBird imagery but no lidar data.
    • The remaining large section of the city which had May QuickBird and lidar data.
  • Each of the 3 rule sets was created using a similar process:
    • Imagery was examined to locate a representative tile.
    • Supportive image layers were created such as Normalized Difference Vegetation Index. (NDVI) and Leefs Sigma Edge Extraction to aid classification efficacy.
    • Image objects were generated representing roads and water features from shapefiles and classified as such.
    • If lidar data were available images were first segmented into tall and short features.
    • Remaining portions of the image were classified utilizing algorithms available in eCognition taking advantage of spectral information as well as other elements of image interpretation such as context, shape, size, site, association, pattern, shadows and texture.
    • Classification was exported from eCognition into a TIF raster file.
  • Each rule set was fine tuned and tested on additional random tiles throughout the city.
  • Each of the final rule sets was used to classify all the tiles comprising its section of St. Paul using eCognition Server.
  • Individual classified tiles were joined into a single mosaic using geometric seam lines in ERDAS Imagine Mosaic Pro.
  • The three different sections of the city, represented by 402 individual tiles, were combined into a single classification file.
  • The resulting classification was used to create an accuracy assessment in ERDAS Imagine using 1,067 stratified random points.
  • The classification mosaic was then manually examined and edited to eliminate errors.
  • Error corrections were re]run in eCognition Server to incorporate corrections.
  • The final land cover mosaic was manipulated by ERDAS Imagine and ArcGIS into the output geodatabase utilizing both raster and vector forms of the data.
  • A Python script was written to summarize classification information into various shapefiles such as parcels and neighborhoods.

An object‐oriented image analysis approach in which the imagery was first segmented into objects with
similar pixels based on the spatial, as well as the spectral‐radiometric (color) attributes was used for the
image classification. Research has shown that it is the best approach for classification of high resolution
imagery (Blaschke, 2010; Platt and Rapoza, 2008). Objects include more information than individual
pixels, enabling the ability to take advantage of all the elements of image interpretation, particularly
spatial information, including shape, size, pattern, texture, and context. Context is especially useful.
Humans intuitively integrate “pixels” into objects and use contextual relationships to interpret images
and draw intelligent inferences from them. Ancillary data such as GIS layers of, for example, streets and
water bodies, could also be incorporated into the decision rules.

The object]oriented image analysis process in eCognition can broadly be split into two components,
segmentation and classification. Segmentation primarily uses spectral information about individual
pixels in the imagery to combine them into larger image objects or segments. As an example, individual
pixels which comprise the roof of, for example, a building are combined where the brightness, NDVI and
color information are similar to form an image object that represents the building. Other scaling
information can be specified to regulate the size range of the desired objects. Once these image objects
are created, they can be classified using a multitude of decision rules which utilize not only their spectral
characteristics but also spatial information such as shape, size, proximity to other object types, texture,
and context. The overall process is dependent on the quality of the initial segmentation into image
objects.

Accuracy assessment was performed on the results after the tiles were edited for corrections. The
accuracy assessment was executed by generating stratified random points across the image and
comparing the classified results to reference imagery (color ortho photos provided by the City and
imagery from ArcGIS online). Stratified random point selection assures each class will be weighted
proportionately to the total number of features in that class across the image. There were 1,067 points
in the sample.

Tabulation of the percent area of each of the six land cover classes.
Class Percent
Tree Canopy 32.5
Grass/Shrub 22.6
Impervious 23.9
Buildings 13.7
Water 7.1
Bare Soil 0.2
Total 100.0

Source Scale Denominator


Section 3 Spatial Data Organization Information Top of full metadata Top of page
Native Data Set Environment eCognition Developer version 8.0, ArcGIS version 9.3.1, and ERDAS Imagine version 2010
Geographic Reference for Tabular Data
Spatial Object Type Raster
Vendor Specific Object Types
Tiling Scheme


Section 4 Spatial Reference Information Top of full metadata Top of page
Horizontal Coordinate Scheme Universal Transverse Mercator
Ellipsoid Geodetic Reference System 80
Horizontal Datum NAD83
Horizontal Units Meters
Distance Resolution 30
Altitude Datum Not applicable
Depth Datum Not applicable
Cell Width
Cell Height
UTM Zone Number 15N


Section 5 Entity and Attribute Information Top of full metadata Top of page
Entity and Attribute Overview The overall accuracy of 90.3 percent and userfs and producerfs accuracies of 93.4 and 91.2 percent for tree cover meets the expected accuracy goals of the project. The primary errors are some confusion
between trees and grass/shrub, between buildings and impervious, and grass and impervious. The
single largest area of confusion was the 17 impervious points which were erroneously classified as grass
and shrub. Many of these occurred at the boundary of tree canopy and roadway where pixel gbleedh
may have contributed to the errors.
Although the pixel size of the pan]sharpened QuickBird imagery is approximately 0.6 meters, the lower
limit for size detection of individual objects is between 2 and 3 meters square. More specifically, to
improve the spatial resolution of our imagery we used a pan]sharpening process which takes QuickBird
spectral information from the 2.4 meter multispectral pixels and distributes it mathematically to the
higher resolution 0.6]meter panchromatic pixels to create 0.6]meter multispectral pixels. Although creating very good multispectral imagery, it is not quite the same if the original multispectral data was 0.6 meter resolution. The is some gbleedingh of the spectral information which made it difficult to isolate narrow areas of impervious cover such as sidewalks and also caused the gfuzzyh boundaries noted in the above paragraph.
Another limitation to the study was the temporal mismatch between the QuickBird and lidar imagery.
The QuickBird image was acquired approximately one year after the lidar data and this resulted in
several inconsistencies in the classification. An example would be trees present in the lidar but
subsequently removed prior to the QuickBird acquisition. Where the tree had overhung a street, the
software interpreted the existing height as an impervious object and classified it as a building. If it
overhung grass, it was interpreted as a tree. Where found, these errors were manually corrected. More
difficult to correct was the reverse situation where a new tree planting does not have matching lidar
height information. Many of these were classified as grass/shrub areas.
As a final note, for this analysis, tree canopy was allowed to grow over any street and street layers and
was not truncated at the edge of the street. The exception to this would be the parcel analyses that
follow. In these cases, only the percentage of the tree canopy that actually falls within the border of the
parcel is included. Since parcels do not extend into the street, the canopy does not as well.
Entity and Attribute Detailed Citation


Section 6 Distribution Information Top of full metadata Top of page
Publisher Remote Sensing and Geospatial Analysis Lab, Univeristy of Minnesota
Publication Date 04/12/2011
Contact Person Information Marvin Bauer, Professor
Remote Sensing and Geospatial Analysis Lab, Univeristy of Minnesota
1530 Cleveland Avenue North
St. Paul , MN 55108
Phone: (612)624-3703
Fax: (612)625-5212
Email: mbauer@umn.edu
Distributor's Data Set Identifier wood_final_classification_x.img
Distribution Liability This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.This data may be used for educational and non-commercial purposes, provided proper attribution is given. Secondary distribution of the data is permitted, but not supported by the University of Minnesota. By accepting the data, the user agrees not to transmit this data or provide access to it or any part of it to another party unless the user includes with the data a copy of this disclaimer.
Transfer Format Name HFA/Erdas Imagine Images (.img)
Transfer Format Version Number
Transfer Size
Ordering Instructions see website or contact infosee website or contact info
Online Linkage Click here to download data. (See Ordering Instructions above for details.) By clicking here, you agree to the notice in "Distribution Liability" in Section 6 of this metadata.


Section 7 Metadata Reference Information Top of full metadata Top of page
Metadata Date 12/05/2006
Contact Person Information Marvin Bauer, Professor
Remote Sensing and Geospatial Analysis Lab, Univeristy of Minnesota
1530 Cleveland Avenue North
St. Paul , MN 55108
Phone: (612)624-3703
Fax: (612)625-5212
Email: mbauer@umn.edu
Metadata Standard Name Minnesota Geographic Metadata Guidelines
Metadata Standard Version 1.2
Metadata Standard Online Linkage http://www.gis.state.mn.us/stds/metadata.htm


This page last updated 10/07/2014 Go back to top