app.py

import cv2
from keras.models import load_model
import gradio as gr
import numpy as np
from tensorflow.keras.preprocessing.image import ImageDataGenerator

model = load_model('Chicken_Gizzard_Updated_model.h5',compile=True)
class_names={0:'Normal Appearance',
             1:'The proventriculusof infected chickens showing several ecchymotic hemorrhages on the tip of the proventricular glandsat 3 dpi',
             2:'edema with increased number of solitary and coalesced ecchymotic hemorrhages on theproventricular glands at 4 dpi',
             3:'and numerous hemorrhagic spots coalesced to form brush paintappearance on the entire mucosa at 5 dpi'}

def Predict_Gizzard(img):
  img = img.reshape((1, img.shape[0], img.shape[1], img.shape[2]))

  # Create the data generator with desired properties
  datagen = ImageDataGenerator(
    rotation_range=30,
    width_shift_range=0.1,
    height_shift_range=0.1,
    shear_range=0.1,
    zoom_range=0.1,
    horizontal_flip=True,
    fill_mode="nearest",
  )
  # Generate a batch of augmented images (contains only the single image)
  augmented_images = datagen.flow(img, batch_size=1)
  # Get the first (and only) augmented image from the batch
  augmented_img = next(augmented_images)[0]
  img=cv2.resize(augmented_img.astype(np.uint8),(224,224))
  class_no=model.predict(img.reshape(1,224,224,3)).argmax()
  name=class_names.get(class_no)
  return name


interface=gr.Interface(fn=Predict_Gizzard,inputs='image',outputs=[gr.components.Textbox(label='Your Result')],
                      examples=[['Class A.PNG'],['Class B.PNG'],['Class C.PNG'],['Class D.PNG']])

interface.launch(debug=True)