predict.py

from keras.preprocessing.image import load_img, img_to_array
import numpy as np
import operator
import json

def preprocess_image(img_path):
    img = load_img(img_path, target_size=(224, 224))
    x = img_to_array(img)
    x = np.expand_dims(x, axis=0)
    x = x/255
    return x

def model_predict(img_path, model):
    predictions = model.predict(img_path)
    pred_5 = np.argsort(predictions)[0][-3:]
    top_5 = {}
    labels_dict = {'Apple Scab': 0, 'Apple Black rot': 1, 'Apple Cedar rust': 2, 'Apple healthy': 3, 'Blueberry healthy': 4, 'Cherry(Sour) Mildew': 5, 'Cherry(Sour) healthy': 6, 'Corn Leaf spot': 7, 'Corn Common rust': 8, 'Corn Northern Leaf Blight': 9, 'Corn healthy': 10, 'Grape Black rot': 11, 'Grape Black measles': 12, 'Grape Leaf Blight': 13, 'Grape healthy': 14, 'Citrus greening': 15, 'Peach Bacterial spot': 16, 'Peach healthy': 17, 'Bell_Pepper Bacterial spot': 18, 'Bell_Pepper 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 Two-spotted spider mite': 33, 'Tomato Target Spot': 34, 'Tomato Yellow Leaf Curl Virus': 35, 'Tomato mosaic virus': 36, 'Tomato healthy': 37}
    for i in pred_5:
        rank = predictions[0][i]
        for kee, val in labels_dict.items():
            if i == val:
                top_5[kee] = round(rank * 100, 3)
    sorted_x2 = sorted(top_5.items(), key=operator.itemgetter(1), reverse=True)
    return sorted_x2

def description(result):
    result2 = dict(result)
    top_hit = list(result2.keys())[0]
    top_val = list(result2.values())[0]
    dis_file = "models/disease_description.csv"
    result_final = {}
    with open(dis_file, 'r') as fh_in:
        for line in fh_in:
            line = line.strip().split(",")
            result_final[line[0]] = line[1]
    result = {}
    for kee, val in result_final.items():
        if top_hit in kee:
            new = top_hit + " : "+ str(top_val) + "%"
            result[new] = val
    final = [(kee, val) for kee, val in result.items()]
    return final