Automated Geo-referencing of Images Dr. Ronald Briggs Yan Li Email: rbriggs@utdallas. edu yanli@utdallas. edu Geo. Spatial Information Sciences The University of Texas at Dallas
Spatial datasets from different sources need to be accurately aligned geographically in order to be viewed or analyzed together.
Image Geo-referencing – To align a raw image with a real world map coordinate system
Typical images & Reference Vector Map Satellite Imagery (e. g. , Google map screenshot) Historical photo Maps Reference Map City of Dallas Street Centerline Arial photo Drawings (e. g. As-Builts)
Geo-Referencing
Manual Geo-Referencing Process Image Reference map v Key step: Manually identify a set of corresponding linked pairs – i. e, Control Points Pairs (CPPs)
Issues v Manually finding CPPs – – – Time Consuming Tedious Sometimes impossible – Must know a priori the approximate location
The impossible: Finding the location An unknown aerial photo ? City of Dallas Street. Centerline. shp 68, 000 street segments
More Problems – Distorted Images Image might be arbitrarily rotated Scaled And, of course, translated away from its real world origin Skewed Which makes it even harder to identify the location
Objectives v Automated geo-referencing solutions are in demand – Automatically find an unknown image’s location anywhere in a city, county, state …. – Identify the right transformation to correct an image’s deformation – A practical solution • • • Error tolerant Accurate result Fast processing speed
Rationale v Release users from tedious manual work v Increase productivity significantly v Handle cases that are impossible to be georeferenced by hand v Support high accuracy, consistency, and stability v Batch processing
Transformations Handled v Raster images can be under any combinations of – Similarity transformations Rotation Translation Uniform Scaling Rotation – Affine transformations Scaling Rotation Translation Skew Translation Similarity transformations + Differential scaling Skew Uniform scaling
Our Solution - TPPM v Street intersections are plentiful and easy to identify v General idea: Automatically search for corresponding CPPs in image and vector map – Topological Point Pattern Matching (TPPM)
Theory behind TPPM under similarity transformations v Shape preserving property: relative distance and angles are invariant θ= θ' OB/OA = OB' /OA' TPPM under affine transformations v Area preserving property: Shape is not preserved, but the ratio of areas is constant Area(OAB)/Area(OA'B' ) = c
Automated Scheme Extract intersection points from a vector map and calculate its TPP (a one time task) Extract intersection points from a raster image and calculate its TPP (topological point pattern) Compare image’s TPP with vector’s TPP to find candidate sets of CPPs Matching verification and optimal result generation Transform and resample Image TPP = Topological Point Pattern CPP = Control Point Pairs
Application Implementation
Automated Geo-referencing Result Total RMS error: 6. 23 for 14 CPPs Image points include false, missing, inaccurate points
Features v Our algorithm and application can handle: – – Very large vector map Affine distorted images Unknown location Errors such as missing, spurious, inaccurate image points, or mismatched points v Fast - Processing time is down to seconds for large geo -referencing area v Achieves small total RMS ( “quality of the match”) v Highly scalable. Requires only a small subset of image points for matching
Future Work v Address CPP identification under higher order transformations (distortions). v Incorporate automated point selection from the image. v Adaptive pattern matching – Urban area vs rural area
Questions? Thank you! Ron Briggs: Yan Li: rbriggs@utdallas. edu yanli@utdallas. edu
Скачать презентацию Automated Geo-referencing of Images Dr Ronald Briggs Yan
c501bde5da46724246ef6e648ee0d039.ppt