Message Boards

GROUPS:

Staff Picks Data Science Earth Science Environmental Science Image Processing Dynamic Interactivity Wolfram Language User Interface Wolfram Cloud

Jessica Shi

GreenUP Green Space Detector

Jessica Shi, Columbia University

Posted 1 year ago

3222 Views

5 Replies

5 Total Likes

Follow this post

enter image description here

LINK to PROJECT: https://green-up.weebly.com

GreenUP Green Space Detector

Jessica Shi, Anwesha Das, Aranyo Ray, and Claire Morton

We recently submitted this project to the Hack the Helyx Hackathon this past weekend. The theme of the competition was “Sustainability”. Our goal was to create a tool that could determine if your community meets sufficient criteria as per internationally agreed upon standards of green space. With this power of information, we hope that this will encourage advocacy towards green space development.

To visit our webpage that links to our Wolfram interactive tools: https://green-up.weebly.com

Generate satellite image

createmap[place_,radius_]:=map=GeoImage[GeoRangeQuantity[radius,"kilometers"],GeoCenterplace]

In[]:=

chicagomap=createmap

Chicago

CITY

,2.5

In[]:=

Out[]=

Next, we wanted to be able to more easily visually the green areas.

Replace green pixels with red

gmap[map_]:=map2=ColorReplace[map,GreenRed,.26]

In[]:=

chicagomap2=gmap[chicagomap]

In[]:=

Out[]=

The following piece will return the total number of green pixels in the satellite image, the ratio between green pixels and total pixels, and the area in kilometers-squared of green.

Count the number of red pixels from the new map

gpixelcalc[map2_]:=greenpixel=Length@PixelValuePositions[map2,Red,.1]

In[]:=

chicagopixel=gpixelcalc[chicagomap2]

In[]:=

55815

Out[]=

Calculate the proportion of green pixels out of the total pixels

pixelratios[greenpixel_,map2_]:=pixelratio=1.0*greenpixel/(Times@@ImageDimensions[map2])

In[]:=

chicagoratio=pixelratios[chicagopixel,chicagomap2]

In[]:=

0.113099

Out[]=

Multiply the proportion by the real-life area (distance squared)

areacalc[pixelratio_,distance_]:=area=Quantity[(2*distance)^2,"kilometers squared"]*pixelratio

In[]:=

chicagoarea=areacalc[chicagoratio,2.5]

In[]:=

2.82748

Out[]=

We also wanted to use this information to give an easy-to-understand letter grading system to indicate how “green” an area is.

Create the grading letter graphics with color

lettergen[{letter_,color_}]:=Rasterize[Overlay[{Graphics[{color,Disk[]},ImageSizeTiny],Style[letter,40]},Alignment->Center],RasterSize200]

In[]:=

lettergen[{"A",Darker@Green}]

In[]:=

Out[]=

Use the proportion and assign it a grade

grade[ratio_]:=Which[0≤ratio<.05,lettergen[{"F",Red}],.05≤ratio<.1,lettergen[{"D",Orange}],.1≤ratio<.15,lettergen[{"C",RGBColor[1.,0.82,0.21]}],.15≤ratio<.2,lettergen[{"B",Darker@Yellow}],.2≤ratio≤1,lettergen[{"A",Darker@Green}]]

In[]:=

grade[chicagoratio]

In[]:=

Out[]=

Create a graphic for the grade scale

Rasterize[Row[Map[lettergen[#]&,{{"A",Darker@Green},{"B",Darker@Yellow},{"C",RGBColor[1.,0.82,0.21]},{"D",Orange},{"F",Red}}]],RasterSize700]

In[]:=

Out[]=

We also discovered the function DistanceTransform which can visualize distance to the background using value. By highlighting the green spaces with white and making everything else black, the function could be used. This image was then made semi-transparent and overlaid on the satellite image.

Create visual indicating distance to green space

gmap2[map_]:=Show[{map,SetAlphaChannel[DistanceTransform@ColorNegate@Binarize[ColorReplace[ColorReplace[map,WhiteBlack,.2],GreenWhite,.26],.7]//ImageAdjust,0.6]}]

In[]:=

gmap2[chicagomap]

In[]:=

Out[]=

Finally, we created the form pages formatting them for cloud deploy that allow the input of a location and a distance. The first one allows the user to type in an address or the name of a city, while the second one utilizes the users location.

Cloud deploy a form allowing an input of location and distance

CloudDeployFormPage{"location""Location","distance""Number"},Grid[{{"Satellite Image","Red Pixels Pinpoint Detected Green Spaces"},{createmap[#location,#distance],gmap[map]},{"Total Number of Green Pixels",gpixelcalc[map2]},{"Ratio of green pixels to total pixels",pixelratios[greenpixel,map2]},{"Estimated area of green space",areacalc[pixelratio,#distance]},{"Total area in map",Quantity[(2*#distance)^2,"kilometers squared"]},{"Greeness grade for this community",grade[pixelratio]},{"Distance from greenspace map (White/Bright areas indicate furthest distance from greenspace)",gmap2[map]}}]&,AppearanceRules<|"Title"

,"Description""In the location box, type in a city name, street address, or other location name. This will be the center point of the satellite image. In the distance box, type a number in kilometers which will be the radius of the area around the center point in the satellite image; it is recommended to choose a value between 1-5, to ensure that the green can be detected. Press submit to view results."|>,PageTheme"White",Permissions"Public"

In[]:=

CloudObject

https://www.wolframcloud.com/obj/57646fc6-ed8e-482c-af19-9bddd525d2f7



Out[]=

Cloud deploy a form asking for permission to use location and allowing an input of current location and distance

CloudDeployFormPage{EmbeddedHTML@StringJoin["<script>function initYourPosition( position ) { document.getElementsByName('yourPosition')[0].value = ''+position.coords.latitude+','+position.coords.longitude;}</script>","<script>navigator.geolocation.getCurrentPosition(initYourPosition)</script>"],"yourPosition""Location","distance""Number"},Grid[{{"Satellite Image","Red Pixels Pinpoint Detected Green Spaces"},{createmap["navigator.geolocation",#distance],gmap[map]},{"Total Number of Green Pixels",gpixelcalc[map2]},{"Ratio of green pixels to total pixels",pixelratios[greenpixel,map2]},{"Estimated area of green space",areacalc[pixelratio,#distance]},{"Total area in map",Quantity[(2*#distance)^2,"kilometers squared"]},{"Greeness grade for this community",grade[pixelratio]},{"Distance from greenspace map (White/Bright areas indicate furthest distance from greenspace)",gmap2[map]}}]&,AppearanceRules<|"Title"

,"Description""Your web browser will prompt you for permission to use your location, coordinates will automatically be inputted into the yourPosition box. In the distance box, type a number in kilometers which will be the radius of the area around the center point in the satellite image; it is recommended to choose a value between 1-5, to ensure that the green can be detected. Press submit to view results."|>,PageTheme"White",CreateUUID["temp/"],Permissions"Public"

In[]:=

CloudObject

https://www.wolframcloud.com/obj/jessicashi605/temp/44996525-a3d6-4a5d-8a35-476b8732c2ca



Out[]=

In the future, we hope to:

Refine our algorithm by also identifying polygons and larger green spaces

◼

Attempt to classify them as private or public green spaces

◼

Modify our letter-grading system such that it is not solely dependent on the ratio of green area to total area, but also takes in factors like size, distance from the user, and the quality of green spaces into account

◼

Check it out! https://green-up.weebly.com

POSTED BY: Jessica Shi

Answer

5 Replies

Sort By:

Douglas Kubler

Posted 8 months ago

A quick Google search shows the definition of Green Space - "an area of grass, trees, or other vegetation set apart for recreational or aesthetic purposes in an otherwise urban environment." California is the "Golden State" because of the golden color of grasslands as the Spring-time green phase dries up. That should qualify as "other vegetation". I hope the software could be more inclusive.

POSTED BY: Douglas Kubler

Answer

Moderation Team

Moderation Team, WOLFRAM

Posted 8 months ago

Congratulations! This post was featured in the Wolfram Blog *Wolfram Community Takes on Current Events: COVID-19 Data, Green Spaces, Homeschool Puzzles and More*. We are looking forward to your future contributions.

POSTED BY: Moderation Team

Answer

Daniel Lichtblau

Daniel Lichtblau, Wolfram Research

Posted 1 year ago

This is quite nice. One suggestion I will offer is to have an option control color-based grading. This could for example allow for a grey scale. Could be useful for people with various forms of color-blindness. Best of luck with the competition.

POSTED BY: Daniel Lichtblau

Answer

Anwesha Das

Anwesha Das, Delhi Public School Kalyanpur

Posted 1 year ago

Wow! This is awesome!

POSTED BY: Anwesha Das

Answer

Moderation Team

Moderation Team, WOLFRAM

Posted 1 year ago

-- you have earned *Featured Contributor Badge* Your exceptional post has been selected for our editorial column *Staff Picks* http://wolfr.am/StaffPicks and Your Profile is now distinguished by a *Featured Contributor Badge* and is displayed on the Featured Contributor Board. Thank you!

POSTED BY: Moderation Team

Answer

Reply to this discussion

Reply Preview

Attachments

Remove Add a file to this post

Follow this discussion

or Discard

Group Abstract

Feedback