app.py

from flask import Flask, make_response, jsonify, request
import json
from keras.models import load_model
from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D
from keras.layers import Activation, Dropout, Flatten, Dense
from keras.preprocessing import image
import uuid
import os
    

import urllib.request

app = Flask(__name__)
model = load_model('./models/epochs20.h5')

def url_to_jpg(uri,file_path):
    urllib.request.urlretrieve(uri,file_path)

@app.route('/', methods = ["POST"])
def predict():
    # load model
    data = request.get_json()
    uri = data['imageUrl']
    print(uri)
    # uri = "https://firebasestorage.googleapis.com/v0/b/pacman-73ff5.appspot.com/o/events%2Fa%40g.com%2Fimage0.279252745985788751234?alt=media&token=eed00af5-5563-4458-91bd-a46e30a9fbe5"
    
    id = str(uuid.uuid1())
    filename = id+".jpg"
    file_path = "images/" + filename
    url_to_jpg(uri,file_path)

    # print("file Saved")
    img_path = file_path
    img = image.load_img(img_path, target_size=(224, 224))

    img_array = image.img_to_array(img)
    new_array = img_array.reshape(1,224,224,3)
    prediction = model.predict(new_array)

    #####################################################
    ####         status = 1 means accident           ####
    ####         status = 0 means noAccident         ####
    #####################################################

    data = {"status" : 0 }
    if(prediction[0] < 0.5) :
        print("Accident")
        data["status"] = 1
    else:
        print("No Accident Yet")

    response = make_response(
                jsonify(
                    data
                ),
                200,
            )
    response.headers["Content-Type"] = "application/json"
    os.remove(file_path)
    # print("file deleted")

    return response

    # return "hey"


if __name__ == '__main__':
    app.run()