Dragonfly diversity informatics to achieve ecological assessments

• Dragonflies and damselflies (Odonata) are freshwater insects that are used as ecological indicators in ecological assessments. Many amateurs post dragonfly images for identification purposes on social media platforms such as Facebook.
• We are creating a pipeline to bring about biodiversity Informatics mediated by social media, an AI-based app for taxonomy, and a checklist app for reporting dragonfly ecological data , involving layman and citizens as stakeholders. With our tools and apps citizens can record dragonfly data that can be used in research, conservation, and education.

Task 1 [Get Data & EDA]

NOTE :- Before running the script make sure that you have logged in facebook developers and done the nescessary procedure for downloading.

1. While running the script change group variable below to the group id of the page from which you want to scrape and also change the token to the token you have generated. After that just put the location of the /folder/(jsonfile name) in location variable where you want to keep your intermediate json file.

#Main function
def main(x,y):
    group = '216593478390203'
    token = 'EAAJaZBipIeNYBAEuW1is0QNSXH7Bb9e9DJQ5z1tT0fIw8FLMmonqGC4Kim0tsDTj14P4SUZB7SbtPi8GQD3qofhHZAy1EUWCAys1ZBc5RAP91sTc79J1bSXnXZCrsZAsJBx1NDGicejW0G3B0spZA8CfmFeZAzsDJToC4XVWAmobI60at73hZB2J5WZAn1ZBFJrnCOYeiYKHFDAcHYuXZBZCwZAKkHsBK4BLC6wYSaoPpcm73xf4hA2ZAF8YP58sVvtNYBQ1QoZD'
    graph = facebook.GraphAPI(token)
    location_path = "data\demox1.json"
    x = datetime.strptime(x,'%d-%m-%Y')
    y = datetime.strptime(y,'%d-%m-%Y')
    field1 =  f'feed.until({y}).limit(400).since({x})'
    field2 ='{full_picture,attachments{description},updated_time}'
    fieldst = field1+field2
    fields = [str(fieldst)]
    profile = graph.get_object(group,fields = fields) 
    with open(location_path,"w") as outfile: #context handler
	        json.dump(profile,outfile,indent=4, separators=(',', ': '))
	        outfile.write('\n')

2. Set location_date path to the txt file in which date is present write there the start date from which you want to start extracting the data and set the months parameter below to a number which will iterate the process for that many months for example lets say you have written 1-1-2012 as a date in txt file after running the script you will get the data between 1-1-2012 and 1-(1+months)-2012

    location_date = 'data\date.txt'
    f = open(location_date, "r")
    date1 = f.read()
    date1f = datetime.strptime(date1,'%d-%m-%Y')
    date1f = date1f+relativedelta(months=+2)
    date1f = date1f.strftime('%d-%m-%Y')
    f.close()    
    main(date1,date1f)
    xtoc()
    f = open("data\date.txt", "w")
    f.write(date1f)
    f.close()

Now you are ready to run the script run it as many times as you want to iterate through the start and end date like if you want data from 2012 to 2013 and you have choosen months as 2 than write 1-1-2012 in date.txt and run the script 6 times.

3. Then we proceed to perform spatio-temporal analysis.

Analysis

Distribution of observation across the country at district-level.

Frequency of various species observed.

District vs no. of observations

Hotspots for Trithemis Aurora.

Focusing on Palakkad district.

Task 2

• Create Dataset by Downloading, Cleaning and Labelling the images.
• Build a Keras Model for Species Identification.

Preview from the app

Task 3

Login Page

Forget Password Page

Home Page

Species List Page

User Selection Page

Final Submit Preview Page

Data Retrieved From Real-Time Firebase