Spaces:
Running
Running
import plantvision | |
import requests | |
from io import BytesIO | |
import pickle as pkl | |
from flask import Flask, render_template, request, session, jsonify, url_for | |
from PIL import Image | |
import os | |
import time | |
import random | |
from pathlib import Path | |
THIS_FOLDER = Path(__file__).parent.resolve() | |
import os | |
os.environ["TRANSFORMERS_CACHE"] = "/tmp/transformers_cache/" | |
app = Flask(__name__) | |
app.secret_key = 'pi-33pp-co-sk-33' | |
app.template_folder = os.path.abspath(f'{THIS_FOLDER}/web/templates') | |
app.static_folder = os.path.abspath(f'{THIS_FOLDER}/web/static') | |
print(app.static_folder) | |
flowerLayers = None | |
leafLayers = None | |
fruitLayers = None | |
def home(): | |
return render_template('index.html') | |
def guess(): | |
global flowerLayers, leafLayers, fruitLayers | |
if request.method == 'POST': | |
print('Thinking...') | |
img = request.files.get('uploaded-image') | |
feature = request.form.get('feature') | |
tensor = plantvision.processImage(img, feature) | |
predictions = plantvision.see(tensor, feature, 6) | |
with open(f'{THIS_FOLDER}/resources/speciesNameToKey.pkl','rb') as f: | |
speciesNameToKey = pkl.load(f) | |
with open(f'{THIS_FOLDER}/resources/speciesNameToVernacular.pkl','rb') as f: | |
speciesNameToVernacular = pkl.load(f) | |
with open(f'{THIS_FOLDER}/resources/{feature}speciesIndexDict.pkl','rb') as f: | |
speciesNameToIndex = pkl.load(f) | |
urls = [] | |
predicted_image_urls = [] | |
for p in predictions: | |
key = speciesNameToKey[p] | |
img = speciesNameToIndex[p] | |
query = '' | |
for i in p.split(' '): | |
query += i | |
query += '+' | |
urls.append(f'https://www.google.com/search?q={query[:-1]}') | |
predicted_image_urls.append(f"https://storage.googleapis.com/bmllc-images-bucket/images/img{img}.jpeg") | |
names = [] | |
for p in predictions: | |
try: | |
names.append(speciesNameToVernacular[p]) | |
except: | |
names.append(p) | |
response = { | |
'names': names, | |
'species': predictions, | |
'predictions': urls, | |
'images': predicted_image_urls | |
} | |
return jsonify(response) | |
if __name__ == '__main__': | |
app.run(port=int(os.environ.get("PORT", 7860)),host='0.0.0.0',debug=True) | |