application/views.py

from flask import Blueprint, request, url_for
from flask import render_template
from werkzeug.utils import secure_filename
from . import app
import os
os.environ["CUDA_VISIBLE_DEVICES"]="-1"
from tensorflow.keras.models import load_model
import numpy as np
import matplotlib.pyplot as plt
from tensorflow.keras.preprocessing.image import load_img, img_to_array
from . import app
plt.switch_backend('agg')
views = Blueprint("views", __name__)
@views.route('/', methods=['GET','POST'])
def home():
    if request.method == "POST":
        file = request.files['file']
        input_img = secure_filename(file.filename)
        file.save(app.config['IMAGE_UPLOADS']+input_img)

        pred=predict_save(input_img)
        return render_template('home.html', pred=pred, input_img=input_img, pred_img='pred_img.png')
    return render_template('home.html')


@views.route('/about')
def about():
    return render_template('about.html')

##############################################
model = load_model(app.config['MODEL'])
class_names = ['Acne and Rosacea', 'Actinic Keratosis Basal Cell Carcinoma and other Malignant Lesions', 'Atopic Dermatitis ',"Exanthems and Drug Eruptions", "Hair Loss Photos Alopecia and other Hair Diseases", "Herpes HPV and other STDs Photos","Nail Fungus and other Nail Disease","Poison Ivy Photos and other Contact Dermatitis","Psoriasis pictures Lichen Planus and related diseases"]


def predict_save(img):
    my_image = load_img(app.config['IMAGE_UPLOADS']+img, target_size=(128, 128))
    my_image = img_to_array(my_image)
    my_image = np.expand_dims(my_image, 0)
    
    out = np.round(model.predict(my_image)[0], 2)
    fig = plt.figure(figsize=(8, 5))
    plt.barh(class_names, 
            [1,1,1,1,1,1,1,1,1], 
            edgecolor='gray',
            linewidth=2,
            color='white',
            height=0.5)
    plt.barh(class_names,
             out, 
             color='lightgray', 
             height=0.5)
    
    for index, value in enumerate(out):
        plt.text(value/2, index, f"{100*value:.2f}%",fontsize=13, fontweight='bold')
        
    plt.xticks([])
    plt.yticks([0, 1, 2, 3, 4, 5, 6, 7,8], labels=class_names, fontweight='bold', fontsize=14)
    name = app.config['IMAGE_UPLOADS']+'pred_img.png'
    fig.savefig(name, bbox_inches='tight')
    
    return class_names[np.argmax(out)]