Skip to content

ECN Infra Demo

ECN Internal Infrastructure Demo
 
In order to collect available digital vector assets of building footprints statewide (data), QA/QC that data, and build derived products from that data, project internal infrastructure is required. A demonstration of integrated components follows:

ecn_bakersfield_demo0

WMS Identify -- LA Bldgs Raw

WMS Identify — LA Bldgs

  • WMS Layer assets are made available via Web Mapping Services. WMS layers carry styling rules, and are intelligently sampled on the server-side to send only necessary information across slower data links. Selected WMS layers also support feature identification. -wikipedia-
  • Server Assets are delivered upon request by Geoserver, in a plurality of formats and under a security framework based on Spring Security.
  • Source Code Ingestion, processing and reporting are driven by code, be it SQL, python, shell script or other. Code, snippets and certain intermediate assets are stored in Git source control.
     
  • Contents

    Blockgroup ID

    Blockgroup IDs

    • Buildings — retrieved from City of Bakersfield as shapefile (SHP)
    • California_20090801 — NAIP imagery stored internally
    • ca_county_shp — Census 2016 TIGER shapefiles
    • worldcities_ca_min — City name as POINT with population count and dynamic styling
    • osm_roads — from Openstreetmap California data snapshot, extracted and styled for visualization (viz)
    • LA_Bldgs_raw — full dataset from LA County, 2008
    • LA_Bldgs_pt — buildings with attributes, stored as POINT for summary spatial statistics
    • costar_mf_pts0 — Costar Multifamily units, supplied by CEC
    • census_2016_uac — Census 2016 TIGER shapefiles, urban areas layer
    • ca_bg_dark — Census 2016 TIGER shapefiles, blockgroups with labeling
    • ne_10m_urban_areas — simplified urban area boundaries from NaturalEarth dataset
       
  • Orthophotos Aerial imagery with the camera oriented straight-down is sometimes called the Bombardier’s View and is a typical method to create orthophotos. Tracing vector assets over orthophotos is common. Every one of the major imagery providers does have alignment issues in various local geographical areas. In the JOSM Openstreetmap editing environment, there are provisions for storing offsets, per local area, per imagery provider (e.g. MapBox Satellite versus BING in Central Europe). However, given current mapping feature density, it is possible for an OSM editor to simply copy the offsets of other features around them, somewhat analogous to driving a car in traffic at the same speed as the other cars, irregardless of the posted speed limit. Imagery alignment does matter and poor alignment can be the source of errors, but modern image analysis benefits from a crowd effect of multiple, stable reference sets in most all areas. These benefits are most quickly exploited by a human operator; using reference data to stabilize offsets in a machine learning environment is TBD.
     

 

naip_08_bkrsfld_align_test1

sangis_nonresidential_base

newport_beach_sample_15sep16


osm_bldgs_bing_align0

osm_bldgs_mapbox_misalign0

align_labldgs_raw_naip_2009
 

NAIP 2016 Images

    —

MapViewer osmb / NAIP 2014 Whole Counties

osmb_NAIP14

MapViewer osmb / NAIP 2014 Detail


    —

DOQQ coverage as of 13Feb17

Calibrating the placement and overlap of DOQQs

Masking Trouble in Paradise, California.


    —

digital ortho quarter quad tiles (DOQQs) * Native Resolution

digital ortho quarter quad tiles (DOQQs) * 3x Resolution


    —

digital ortho quarter quad tiles (DOQQs) * ex0

digital ortho quarter quad tiles (DOQQs) * Infrared Band


    —

An Infrared (IR) View of Tree-filled Downtown Arcata, Calif.

DOQQ Infrared (IR) areas processed as of 13Feb17