geeViz Google Maps Platform Examples

This notebook demonstrates geeViz.googleMapsLib — 24 functions for ground-truthing remote sensing analysis using Google Maps Platform APIs.

Requirements:

• pip install geeViz

• A .env file in the geeViz directory with:

  GEMINI_API_KEY=your_key_here
  GOOGLE_MAPS_PLATFORM_API_KEY=your_key_here
  

• For segmentation: pip install geeViz[segmentation]

Section	APIs Used
1. Geocoding	Geocoding, Reverse Geocoding, Address Validation
2. Street View	Street View Static, Panorama Stitching
3. Places	Places (New), Place Photos
4. Environment	Elevation, Air Quality, Solar, Timezone
5. AI Analysis	Gemini Interpretation, Object Detection
6. Segmentation	SegFormer Semantic Segmentation

import geeViz.googleMapsLib as gm
from IPython.display import display, HTML, Image as IPImage, Markdown
import json, os, time

output_dir = os.path.join(os.path.dirname(os.getcwd()), "examples", "outputs", "google_maps")
os.makedirs(output_dir, exist_ok=True)
print(f"Output: {output_dir}")
print(f"Maps key: {'set' if gm._get_api_key() else 'MISSING'}")

---

1. Geocoding

Convert addresses to coordinates and back.

# Forward geocode
loc = gm.geocode("100 S State St, Salt Lake City, UT")
print(f"Address: {loc['formatted_address']}")
print(f"Coords: {loc['lat']}, {loc['lon']}")
print(f"Accuracy: {loc['location_type']}")

# Reverse geocode
rev = gm.reverse_geocode(loc['lon'], loc['lat'])
print(f"\nReverse: {rev['formatted_address']}")

# Address validation
val = gm.validate_address("4200 olympic way slc ut")
if val:
    print(f"\nValidated: {val['formatted_address']}")
    print(f"Complete: {val['verdict']['address_complete']}")
    if val.get('usps_data'):
        print(f"USPS: {val['usps_data']['standardized_address'].get('firstAddressLine')}")

---

2. Street View

Fetch Street View imagery — single frames, cardinal directions, stitched panoramas.

# Check coverage
meta = gm.streetview_metadata(loc['lon'], loc['lat'])
print(f"Status: {meta['status']}, Date: {meta.get('date')}")

# Single image
img = gm.streetview_image(loc['lon'], loc['lat'], heading=120)
if img:
    display(IPImage(data=img, width=600))
    print(f"Single frame: {len(img):,} bytes")

# 360° stitched panorama
pano = gm.streetview_panorama(loc['lon'], loc['lat'], heading=120, fov=360)
if pano:
    display(IPImage(data=pano, width=900))
    print(f"360° panorama: {len(pano):,} bytes")

# HTML panel with cardinal directions
html = gm.streetview_html(loc['lon'], loc['lat'])
if html:
    display(HTML(html))

---

3. Places

Search for places near a location.

# Text search
places = gm.search_places("coffee shop", lat=loc['lat'], lon=loc['lon'], radius=2000)
print(f"Found {len(places)} coffee shops:\n")
for p in places[:5]:
    rating = f" ({p['rating']}★)" if p.get('rating') else ""
    print(f"  {p['display_name']}{rating}: {p['address']}")

# Nearby search
parks = gm.search_nearby(loc['lat'], loc['lon'], radius=3000, included_types=["park"])
print(f"\nFound {len(parks)} parks nearby")
for p in parks[:3]:
    print(f"  {p['display_name']}: {p['address']}")

---

4. Environment

Elevation, air quality, solar potential, and timezone.

# Elevation
elev = gm.get_elevation(loc['lon'], loc['lat'])
print(f"Elevation: {elev:.0f} m ({elev * 3.28084:.0f} ft)")

# Elevation profile
profile = gm.get_elevation_along_path(
    [(-111.89, 40.77), (-111.80, 40.68)],  # SLC downtown to Wasatch foothills
    samples=10,
)
print(f"\nElevation profile ({len(profile)} points):")
for p in profile:
    print(f"  ({p['lat']:.4f}, {p['lon']:.4f}): {p['elevation']:.0f} m")

# Air quality
aq = gm.get_air_quality(loc['lon'], loc['lat'])
if aq:
    print(f"AQI: {aq['aqi']} — {aq['category']}")
    print(f"Dominant pollutant: {aq['dominant_pollutant']}")

# Solar potential
solar = gm.get_solar_insights(loc['lon'], loc['lat'])
if solar:
    print(f"\nSolar potential:")
    print(f"  Max panels: {solar['max_panels']}")
    print(f"  Annual output: {solar['max_annual_kwh']:,.0f} kWh")
    print(f"  Roof area: {solar['roof_area_m2']:.0f} m²")
# Show static map of the location
static_map = gm.get_static_map(loc['lon'], loc['lat'],zoom=16,markers=[(loc['lon'],loc['lat'])])
if static_map:
    display(IPImage(data=static_map, width=600))

# Timezone
tz = gm.get_timezone(loc['lon'], loc['lat'])
if tz:
    print(f"\nTimezone: {tz['timezone_id']} ({tz['timezone_name']})")

---

5. AI Analysis (Gemini)

Use Gemini to interpret and label Street View imagery. Requires GEMINI_API_KEY.

display(IPImage(data=pano, width=900))
print('Intpretting image...')
# Interpret a panorama
result = gm.interpret_image(
    pano,
    context=f"{loc['formatted_address']}. Downtown Salt Lake City.",
)
print('Finished intpretting image')

display(Markdown(result['description']))
display(Markdown(result['object_counts']))

# Label objects with bounding boxes
labeled = gm.label_streetview(
    loc['lon'], loc['lat'],
    heading=120, fov=360,
    max_labels=20,
)
if labeled:
    display(IPImage(data=labeled['image'], width=900))
    print(f"Detected {len(labeled['detections'])} objects")
    display(Markdown(labeled['summary']))

---

6. Semantic Segmentation (SegFormer)

Pixel-level land cover classification. Requires pip install geeViz[segmentation].

# Segment the panorama with SegFormer B4
try:
    t0 = time.time()
    seg = gm.segment_image(pano, model_variant="b4", broad_categories=False)
    print(f"Segmented in {time.time()-t0:.1f}s")
    display(IPImage(data=seg['colored_image'], width=900))
    display(Markdown(seg['summary']))
except ImportError:
    print("SegFormer requires: pip install geeViz[segmentation]")

# Segment with broad land cover categories
try:
    seg_broad = gm.segment_image(pano, model_variant="b4", broad_categories=True)
    display(IPImage(data=seg_broad['colored_image'], width=900))
    display(Markdown(seg_broad['summary']))
except ImportError:
    print("SegFormer requires: pip install geeViz[segmentation]")

---

7. Roads & Static Maps

# Nearest roads
roads = gm.nearest_roads(loc['lon'], loc['lat'])
print(f"Nearest roads: {len(roads)} segments")
for r in roads:
    print(f"  ({r['lat']:.5f}, {r['lon']:.5f})")

# Static map
static = gm.get_static_map(
    loc['lon'], loc['lat'],
    zoom=16, size="640x400", maptype="hybrid",
    markers=[(loc['lon'], loc['lat'])],
)
if static:
    display(IPImage(data=static, width=640))
    print(f"Static map: {len(static):,} bytes")

---

Summary

Category	Functions	API Key
Geocoding	geocode, reverse_geocode, validate_address	GOOGLE_MAPS_PLATFORM_API_KEY
Places	search_places, search_nearby, get_place_photo	GOOGLE_MAPS_PLATFORM_API_KEY
Street View	streetview_metadata, streetview_image, streetview_panorama, streetview_html	GOOGLE_MAPS_PLATFORM_API_KEY
AI Analysis	interpret_image, label_streetview	GEMINI_API_KEY
Segmentation	segment_image, segment_streetview	None (local model)
Elevation	get_elevation, get_elevations, get_elevation_along_path	GOOGLE_MAPS_PLATFORM_API_KEY
Environment	get_air_quality, get_solar_insights, get_timezone	GOOGLE_MAPS_PLATFORM_API_KEY
Maps & Roads	get_static_map, snap_to_roads, nearest_roads	GOOGLE_MAPS_PLATFORM_API_KEY