Frank Falzone Ross Meyer FR 3262 10December2012 Outline Hybrid hazels in context Desired project output Data acquisition Procedures Accuracy assessment Results Limitations and potential revisions ID: 565684
Download Presentation The PPT/PDF document "Identifying Potential Plantation Sites f..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Identifying Potential Plantation Sites for Hybrid Hazelnut Production in a Micro-Cooperative Context
Frank
Falzone
Ross Meyer
FR 3262
10.December.2012Slide2
OutlineHybrid hazels in context
Desired project output
Data acquisition
Procedures
Accuracy assessment
Results
Limitations and potential revisionsSlide3
Hybridization trials
Hybrid Hazels: The Future is Nuts
Current industry and EFB
Current state of
midwest
industrySlide4
Potential Plantation Sites
Currently underutilized land
Along regularly driven routes
Semi-rural residential or hobby farm
Away from road salt zone
At least 1/20
th acreSlide5
Procedure Outline - Data
Download shapefiles of streets, county
roads, major
roads
, and
parcels of Hennepin County
Clip shapefiles to general AOI
PLSS sections containing regularly traveled roadsSlide6
Area of interest:Portions of Maple Grove, Corcoran,
Rogers, St. Michael, Hanover, MNSlide7
Pre-processingDissolve desired sections into 1 polygon
Begin selecting and exporting features not desired
Parcels less than 1.5 acres
Roads with a 55, 75, or 150 foot
buffer, depending on road size
Erase these undesired areas Slide8Slide9Slide10
Pre-processing continued
Convert single polygon shapefile into multi-polygon
Delete small
polygon fragmentsSlide11
Pre-processingDownload 4 band
CIR
imagery for Hennepin County 2008 from
MNGeo
(NAIP)
Contained information for the small
portion
of our AOI in Wright County
Summer aerial photographsOne meter spatial resolutionSlide12
Classification
Open in ERDAS to run supervised classification
Classes:
Grass
Unused pasture
Agriculture
Developed
Open waterForestSlide13
Classified Image SampleSlide14
Accuracy Assessment
Number of Pixels Classified as -
Om.
Prod.
Ref. Class
Forest
Ag
Developed
Unused
Grass
Total
Error
Acc. (%)
Forest
9
2
0
1
0
12
25
75
Ag
3
11
0
7
0
21
47.62
52.38
Developed0030030100Unused0104052080Grass01053966.6733.33Total1215317350nanaCo. Error2526.67076.470nananaUser Acc.%7573.3310023.53100nana60
60% overall Accuracy with Grass and Unused Pasture as individual classesSlide15
Accuracy Assessment Continued
Number of Pixels Classified as -
Om.
Prod.
Ref. Class
Forest
Ag
Developed
Unused/Grass
Total
Error (%)
Acc. (%)
Forest
9
2
0
1
12
25
75
Ag
3
11
0
7
21
47.62
52.38
Developed
0
0
3030100Unused/Grass020121414.2985.71Total121532050nanaCo. Error (%)2526.67040nananaUser Acc. %7573.3310060nana7070% Overall Accuracy with Unused Pasture and Grass as a combined classSlide16
Results
Total classified acres of suitable land:
1383 grass
1983
unused pastureSlide17
ResultsSlide18
ResultsSlide19
Limitations / Possible Improvements
Classification
Use spring/fall imagery for certain classes
More training sites
Better definition of desired classes
Area of interest filtering
Split grass/unused pasture polygons along parcel lines and remove those patches less than 1/20
th acre in a single parcelSlide20
Questions?