finalized image processing model

.gitattributes

+*.hdf5 filter=lfs diff=lfs merge=lfs -text

Help.txt

+1. Training.ipynb file contain the training code which i used to train the model first. I used sequential model to train our model using new Plant leaves disease dataset.
+2. After training of our model is done I used that trained file in my GUI.
+3. Then I used keras model to Predict the results according to our trained model.
+4. The trained system can only pridict the diseases of the plant on which it is trained on.
+
+
+**************************
+
+
+1. I have an image processing model which detect dieseased plants according to the leaf.
+     Trained over many data ( 1gb ), i will provide you the model and the enviorment files.
+    The model should be added to my flask back end application as a function.
+
+2. The flask backend application code will also be given to you, the database which is used is mongoDB.
+
+3. Host the flask application on heroku.( or locally ).
+
+4. From mongoDB take images and process through the python image processing model( now in flask app), 
+    and give the result in the backend as diseased or not
\ No newline at end of file

__init__.py

+import json
+from flask import Flask, request, render_template, json
+from flask_pymongo import PyMongo
+from PIL import Image
+from h5py._hl import datatype
+from tensorflow.keras.models import load_model
+from tensorflow.keras.preprocessing import image
+import numpy as np
+import pandas as pd
+import base64
+
+
+model = load_model('train_data.hdf5')
+model.load_weights('weights.hdf5')
+
+
+# This function will return the label for the given image
+def prediction(img_path):
+    img = image.load_img(img_path, target_size=(224, 224))
+    img = image.img_to_array(img)
+    img = np.expand_dims(img, axis=0)
+    img = img / 255
+    pred = model.predict(img)
+    y = np.argmax(pred)
+    classname = pd.read_csv('classes.csv', encoding='latin1')
+    result = classname.iloc[y][1:3]
+    return result.to_json()
+
+# initializing the flask and mongodb instance
+app = Flask(__name__)
+
+
+app.config["MONGO_URI"] = "mongodb+srv://flask-db:agri1234567@afcluster0.uyijz.mongodb.net/flask-backend?retryWrites=true&w=majority"
+mongo = PyMongo(app)
+
+# Currenlty I am uploading image thorough form request.
+@app.route('/')
+def index():
+    return render_template('index.html')
+
+
+# This route shall accept the file from your react native app to be uploaded to the server and it wll add that image to the database and
+# the uploads folder where all the images will be stored
+@app.route('/add', methods=['GET', 'POST'])
+def add():
+    if request.method == 'POST':
+        data = request.form['file']
+        # wriet the data to a file
+
+        with open('encoded.txt', 'w') as f:
+            f.write(data)
+        file = open('encoded.txt', 'rb')
+        byte = file.read()
+        file.close()
+
+        decodeit = open('uploaded_image.jpg', 'wb')
+        decodeit.write(base64.b64decode((byte)))
+        decodeit.close()
+        imageclass = prediction('uploaded_image.jpg')
+        mongo.db.image.insert({'db_image_name': data, 'image_class': imageclass})
+        data2 = json.loads(imageclass)
+        return "Predicted Disease is " + data2['classes'] + " and Recommended Pesticide is " + data2['pesticide']
+    else:
+        return "Error"
+
+
+if __name__ == "__main__":
+  app.run(host='0.0.0.0', port=5000)

app.py

+import json
+from flask import Flask, request, render_template, json
+#from flask_pymongo import PyMongo
+from PIL import Image
+from h5py._hl import datatype
+from tensorflow.keras.models import load_model
+from tensorflow.keras.preprocessing import image
+import numpy as np
+import pandas as pd
+import base64
+
+model = load_model('train_data.hdf5')
+model.load_weights('weights.hdf5')
+
+
+# This function will return the label for the given image
+def prediction(img_path):
+    img = image.load_img(img_path, target_size=(224, 224))
+    img = image.img_to_array(img)
+    img = np.expand_dims(img, axis=0)
+    img = img / 255
+    pred = model.predict(img)
+    y = np.argmax(pred)
+    classname = pd.read_csv('classes.csv', encoding='latin1')
+    result = classname.iloc[y][1:3]
+    return result.to_json()
+    #return "Predicted Disease is " + str(diseaseName) + " and Recommended Pesticide is " + str(pesticidename)
+
+
+
+# initializing the flask and mongodb instance
+app = Flask(__name__)
+
+
+# app.config["MONGO_URI"] = "mongodb://localhost:27017/leafdiseasedetection"
+# mongo = PyMongo(app)
+
+# Currenlty I am uploading image thorough form request.
+@app.route('/')
+def index():
+    return render_template('index.html')
+
+
+# This route shall accept the file from your react native app to be uploaded to the server and it wll add that image to the database and
+# the uploads folder where all the images will be stored
+@app.route('/add', methods=['GET', 'POST'])
+def add():
+    if request.method == 'POST':
+        data = request.form['file']
+        # wriet the data to a file
+
+        with open('encoded.txt', 'w') as f:
+            f.write(data)
+        file = open('encoded.txt', 'rb')
+        byte = file.read()
+        file.close()
+
+        decodeit = open('uploaded_image.jpg', 'wb')
+        decodeit.write(base64.b64decode((byte)))
+        decodeit.close()
+        imageclass = prediction('uploaded_image.jpg')
+        # mongo.db.image.insert({'db_image_name': data, 'image_class': imageclass})
+        data = json.loads(imageclass)
+        return "Predicted Disease is " + data['classes'] + " and Recommended Pesticide is " + data['pesticide']
+    else:
+        return "Error"
+
+
+if __name__ == "__main__":
+    app.run()

class.csv

+,label
+Apple___Apple_scab,0
+Apple___Black_rot,1
+Apple___Cedar_apple_rust,2
+Apple___healthy,3
+Blueberry___healthy,4
+Cherry_(including_sour)___Powdery_mildew,5
+Cherry_(including_sour)___healthy,6
+Corn_(maize)___Cercospora_leaf_spot Gray_leaf_spot,7
+Corn_(maize)___Common_rust_,8
+Corn_(maize)___Northern_Leaf_Blight,9
+Corn_(maize)___healthy,10
+Grape___Black_rot,11
+Grape___Esca_(Black_Measles),12
+Grape___Leaf_blight_(Isariopsis_Leaf_Spot),13
+Grape___healthy,14
+Orange___Haunglongbing_(Citrus_greening),15
+Peach___Bacterial_spot,16
+Peach___healthy,17
+"Pepper,_bell___Bacterial_spot",18
+"Pepper,_bell___healthy",19
+Potato___Early_blight,20
+Potato___Late_blight,21
+Potato___healthy,22
+Raspberry___healthy,23
+Soybean___healthy,24
+Squash___Powdery_mildew,25
+Strawberry___Leaf_scorch,26
+Strawberry___healthy,27
+Tomato___Bacterial_spot,28
+Tomato___Early_blight,29
+Tomato___Late_blight,30
+Tomato___Leaf_Mold,31
+Tomato___Septoria_leaf_spot,32
+Tomato___Spider_mites Two-spotted_spider_mite,33
+Tomato___Target_Spot,34
+Tomato___Tomato_Yellow_Leaf_Curl_Virus,35
+Tomato___Tomato_mosaic_virus,36
+Tomato___healthy,37

classes.csv

+label,pesticide,classes
+0,Tebuconazole,Apple Apple scab
+1,Myclobutanil,Apple Black rot
+2,Myclobutanil (Immunox),Apple Cedar apple rust
+3,,Apple healthy
+4,,Blueberry healthy
+5,Merivon fluxapyroxad+pyraclostrobin,Cherry (including sour) Powdery mildew
+6,,Cherry (including sour) healthy
+7,Patch Pro.,Corn (maize) Cercospora leaf spot Gray leaf spot
+8,Common Fungicides,Corn (maize) Common rust 
+9," 
+Shamrock Al-sach With Logo Label Mark M45 Mancozeb ",Corn (maize) Northern Leaf Blight
+10,,Corn (maize) healthy
+11,"Mancozeb, and Ziram",Grape Black rot
+12,"Mancozeb, and Ziram",Grape Esca (Black Measles)
+13,"Bordeaux mixture (1.0%),Mancozeb (0.2%), Topsin-M (0.1%), Ziram (0.35%) or
+Captan (0.2%)",Grape Leaf blight (Isariopsis Leaf Spot)
+14,,Grape healthy
+15,"No cure, Redo Process",Orange Haunglongbing (Citrus greening)
+16,"Copper, oxytetracycline",Peach Bacterial spot
+17,,Peach healthy
+18,Copper and Mancozeb sprays,Pepper   bell Bacterial spot
+19,,Pepper   bell healthy
+20,Avoid Overhead Irrigation and allow sufficient aeration,Potato Early blight
+21,prophylactic spray ofmancozeb,Potato Late blight
+22,,Potato healthy
+23,,Raspberry healthy
+24,,Soybean healthy
+25,Apply sulfur or copper-based fungicides,Squash Powdery mildew
+26,Copper-Count-N at 1.5 to 2 quart/100 gal water. Group M1 fungicide,Strawberry Leaf scorch
+27,,Strawberry healthy
+28,Bonide Liquid Copper Fungicide concentrate or Bonide Tomato & Vegetable,Tomato Bacterial spot
+29,Bonide Liquid Copper Fungicide concentrate or Bonide Tomato & Vegetable,Tomato Early blight
+30,Bonide Liquid Copper Fungicide concentrate or Bonide Tomato & Vegetable,Tomato Late blight
+31,Bonide Liquid Copper Fungicide concentrate or Bonide Tomato & Vegetable,Tomato Leaf Mold
+32,Bonide Liquid Copper Fungicide concentrate or Bonide Tomato & Vegetable,Tomato Septoria leaf spot
+33,bifenthrin and permethrin,Tomato Spider mites Two-spotted spider mite
+34,bifenthrin and permethrin,Tomato Target Spot
+35,Imidacloprid,Tomato Tomato Yellow Leaf Curl Virus
+36,"Viral Infection, Remove all infected Plants",Tomato Tomato mosaic virus
+37,,Tomato healthy

requirements.txt

+numpy~=1.19.2
+pandas~=1.3.3
+flask~=1.1.2
+pillow~=8.3.1
+tensorflow~=2.6.0
+gunicorn~=20.1.0
\ No newline at end of file

temp.py

+import os
+import pickle
+import predict
+import numpy as np
+
+Apple_diseases=['Apple___Apple_scab','Apple___Black_rot','Apple___Cedar_apple_rust','Apple___healthy']
+Badam_diseases=['bad','good']
+Blueberry_diseases=['Blueberry___healthy']
+Cherry_diseases=['Cherry_(including_sour)___healthy','Cherry_(including_sour)___Powdery_mildew']
+Corn_diseases=['Corn_(maize)___Cercospora_leaf_spot Gray_leaf_spot','Corn_(maize)___Common_rust_','Corn_(maize)___healthy','Corn_(maize)___Northern_Leaf_Blight']
+Grape_diseases=['Grape___Black_rot','Grape___Esca_(Black_Measles)','Grape___healthy','Grape___Leaf_blight_(Isariopsis_Leaf_Spot)']
+Orange_diseases=['Orange___Haunglongbing_(Citrus_greening)']
+Paddy_diseases=['Bacterial leaf blight','Brown spot','Leaf smut']
+Peach_diseases=['Peach___Bacterial_spot','Peach___healthy']
+Pepper_diseases=['Pepper,_bell___Bacterial_spot','Pepper,_bell___healthy']
+Potato_diseases=['Potato___Early_blight','Potato___healthy','Potato___Late_blight']
+Raspberry_diseases=['Raspberry___healthy']
+Rose_diseases=['anthracnose','mildew','rust','spot']
+Soyabean_diseases=['Soybean___healthy']
+Squash_diseases=['Squash___Powdery_mildew']
+Strawberry_diseases=['Strawberry___healthy','Strawberry___Leaf_scorch']
+Sunflower_diseases=['blight','mildew','rust']
+Tomato_diseases=['Tomato___Bacterial_spot','Tomato___Early_blight','Tomato___healthy','Tomato___Late_blight','Tomato___Leaf_Mold','Tomato___Septoria_leaf_spot','Tomato___Spider_mites Two-spotted_spider_mite','Tomato___Target_Spot','Tomato___Tomato_mosaic_virus','Tomato___Tomato_Yellow_Leaf_Curl_Virus']
+
+diseases = [Apple_diseases,Badam_diseases,Blueberry_diseases,Cherry_diseases,Corn_diseases,Grape_diseases,Orange_diseases,Paddy_diseases,Peach_diseases,Pepper_diseases,Potato_diseases,Raspberry_diseases,Rose_diseases,Soyabean_diseases,Squash_diseases,Strawberry_diseases,Sunflower_diseases,Tomato_diseases]
+Species = ["Apple","Badam","Blueberry","Cherry","Corn","Grape","Orange","Paddy","Peach","Pepper","Potato","Raspberry","Rose","Soyabean","Squash","Strawberry","Sunflower","Tomato"]
+
+
+def classify_disease(imgPath, species='unknown'):
+  if species=='unknown':
+     species = classify_species(imgPath)[0]
+  #open the list of inference models and build respective modelPath
+  pickle_in = open(model_detail,"rb")
+  models_dict = pickle.load(pickle_in)
+  modelPath = "trainedModels/inferenceModels/protobuf/diseases/"+species+'/'+models_dict[species][0]
+  #the largest probability is provided for the class in prediction
+  prediction = predict.predict(imgPath, modelPath)
+  pred = np.argmax(prediction) 
+  pred_class = diseases[models_dict[species][2]][pred]
+  return [pred_class, str(max(prediction))]

temp2.py

+from PIL import Image
+from tensorflow.keras.models import load_model
+from tensorflow.keras.preprocessing import image
+import numpy as np
+import pandas as pd
+import base64
+
+model = load_model('train_data.hdf5')
+model.load_weights('weights.hdf5')
+
+def prediction(img_path):
+    # PIL image loading
+    img = image.load_img(img_path, target_size=(224, 224))
+    img = image.img_to_array(img)
+    img = np.expand_dims(img, axis=0)
+    img = img/255
+    pred = model.predict(img)
+    y = np.argmax(pred)
+    print(y)
+    classes = pd.read_csv('classes.csv', encoding='latin1')
+    prediction = classes.iloc[1][1:3]
+    return prediction.to_json()
+
+
+# file = open('encoded.txt', 'rb')
+# byte = file.read()
+# file.close()
+
+# decodeit = open('uploaded_image.jpg', 'wb')
+# decodeit.write(base64.b64decode((byte)))
+# decodeit.close()
+
+    
+
+print(prediction('images/blackrot.jpg'))
\ No newline at end of file

templates/index.html

+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta http-equiv="X-UA-Compatible" content="IE=edge">
+    <meta name="viewport" content="width=, initial-scale=1.0">
+    <title>Leaf Disease Prediction</title>
+</head>
+<body>
+    Please download the app to get prediction
+</body>
+</html>
\ No newline at end of file