Transcript Land Cover Mapping Project

Land Cover Classification System
•
•
•
•
•
•
•
•
•
PRESENTATION FLOW
Introduction
Steps of Pre-processing
Image Classification in MadCat
Quality Assurance of Data
Edge Matching
Topology Check
Final Quality Check
Layout Development
Results Validation through Field Survey
INTRODUCTION
Land Cover Mapping is an going project in collaboration
with FAO UN using the technique of Land Cover
Classification System (LCCS) – an important component
of FAO / GLCN approach to create a harmonized and
extensive representation of land cover features
FINAL MOSAICS SCHEMA OF AOI
STEPS OF PRE-PROCESSING
KEY POINTS
• DATA USED
– SPOT 5 meter pansharpened data
• 2.5meter rescaled at 5
meter
• Development of
Working areas
– Working areas of whole
province developed with
minimum overlap
KEY POINTS
• Creation of Subsets within working areas
Segmentation created separately on each subset
SEGMENTATION
• Re-projected into world Mercator (WGS-1984) before
segmentation
• Segmentation scale used 50 (± 15 depending upon the features on
image), shape = 0.1, compactness= 0.9)
polygons from outside the
boundary were deleted
Land Cover Classification Map – Hyderabad & Surroundings
Indus River
Hyderabad
.
Sindh
QUALITY ASSURANCE OF DATA
• Quality Checking – Marking of Errors
• Error Removing
• Re – QC
GENERAL ERROR MARKING
ERROR REMOVING
EDGE MATCHING
TOPOLOGY CHECK
• Removal of Gaps
• Removal of Overlaps
TOPOLOGY CHECK
1. Data was first checked for gap errors
2. Correction of gap errors
3. Then, overlap topology was checked
4. Overlap errors were corrected in MadCat
–
Gap corrected layer was imported in segment layer
–
Polygon containing errors were exported to training areas
–
Then again move back to segment layer by using
Topology over
REMOVAL OF GAPS
REMOVAL OF OVERLAPING ERRORS
Land Cover Mapping
The possible areas of application
would be Agriculture, Forestry,
Environment, Irrigation, Disasters &
Hazards Monitoring, Planning &
Development, Oil & Gas Exploration,
Mining, Wild Life and other
emergent requirements.
RESULTS VALIDATION THROUGH FIELD SURVEY
ROUTE OF SURVEY
SURVEY TEAM
RESULTS VALIDATION THROUGH FIELD SURVEY
WLBA
TCIr
RESULTS VALIDATION THROUGH FIELD SURVEY
TCIr / SaD
SaD / WB
RESULTS VALIDATION THROUGH FIELD SURVEY
SCIr
WB
RESULTS VALIDATION THROUGH FIELD SURVEY
HCIrS
SaD
SCIr – Shrubs Crop Irrigated
HCIr – Herbaceous Crop Irrigated
TCIr – Trees Crop Irrigated
HCRf - Herbaceous Crop Rainfed
WLBA - Water Logged Bare Area
WB – Water Bodies
SaFP – Desert Flat Plain
BL – Barren Land
Thank you