app.py

import numpy as np
import tempfile
import pandas as pd
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout, BatchNormalization, GlobalAveragePooling2D
from tensorflow.keras.models import Model, load_model, Sequential
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.metrics import Precision, Recall
from tensorflow.keras.callbacks import EarlyStopping
from sklearn.utils.class_weight import compute_class_weight
from sklearn.model_selection import train_test_split
from tensorflow.image import resize
import cv2

from tensorflow.keras.utils import to_categorical
import matplotlib.pyplot as plt
import warnings
import warnings
warnings.filterwarnings("ignore")

# print ('modules loaded')


import streamlit as st
import pandas as pd 
import numpy as np
from PIL import Image
import tensorflow.keras as keras

st.title("Skin Cancer Classification App")

models = {
  "Le_Net": load_model('LeNet_5.h5'),
  "Simple_CNN": load_model('Simple CNN.h5'),
  "Alex_Net": load_model('AlexNet.h5'),
  "Deeper_CNN": load_model('Deeper CNN.h5')
}

# Allow user to select model
model_name = st.selectbox("Choose a model", list(models.keys()))
model = models[model_name]

# Upload Image

file = st.file_uploader("Upload an image", type=["png", "jpg", "jpeg"])
# file ='hmnist_28_28_RGB.csv'

print(file)

true_file=pd.read_csv("HAM10000_metadata.csv")


# true_file.apply(lambda x: x["image_id"] ==file)

# uploaded_file = st.file_uploader("Choose a image file", type="jpg")

# if uploaded_file is not None:
#     # Convert the file to an opencv image.
#     file_bytes = np.asarray(bytearray(uploaded_file.read()), dtype=np.uint8)
#     opencv_image = cv2.imdecode(file_bytes, 1)

#     st.image(opencv_image, channels="BGR")
    

classes = {4: ('nv', ' melanocytic nevi'), 6: ('mel', 'melanoma'),
           2 :('bkl', 'benign keratosis-like lesions'), 1:('bcc' , ' basal cell carcinoma'),
           5: ('vasc', ' pyogenic granulomas and hemorrhage'),
           0: ('akiec', 'Actinic keratoses and intraepithelial carcinomae'),
           3: ('df', 'dermatofibroma')}


if file is not None:

    file_bytes = np.asarray(bytearray(file.read()), dtype=np.uint8)
    opencv_image = cv2.imdecode(file_bytes, 1)
    
    
    # temp_dir = tempfile.TemporaryDirectory()
    # temp_file_path = temp_dir.name + "/" + file.name
    
    # # Save the uploaded file to the temporary directory
    # with open(temp_file_path, "wb") as f:
    #     f.write(file.read())
        
    # img = cv2.imread(file)
    # cv2_imshow(img)
    img1 = cv2.resize(opencv_image, (32, 32))
    result = model.predict(img1.reshape(1, 32, 32, 3))
    max_prob = max(result[0])
    class_ind = list(result[0]).index(max_prob)
    class_name = classes[class_ind]
    # print(class_name)
    # count+=1
    # if count>10:
    #     break

        
    # df = pd.read_csv(file)  
    # # Get first row 
    # img_reshaped = image_resize(df)
    
    # # Get prediction
    # pred = model.predict(img_reshaped)
    # label = np.argmax(pred)

    
    # Display image and result 
    col1, col2 = st.columns(2)
    with col1:
        st.header("Input Image")
        st.image(opencv_image, channels="BGR")
    with col2:
        st.header("Results")
        st.write("True Label",true_file[true_file["image_id"]==file]["dx"][0]])
        st.write("Prediction",class_name[0])
        st.metric("Category:", class_name[1])

        





# from google.colab.patches import cv2_imshow
# srcdir = '/kaggle/input/skin-cancer-mnist-ham10000/HAM10000_images_part_1'
# count=0
# for temp in os.listdir(srcdir):
#     img = cv2.imread(os.path.join(srcdir, temp))
#     cv2.imwrite(temp, img)
#     cv2_imshow(img)
#     img = cv2.resize(img, (28, 28))
#     result = model.predict(img.reshape(1, 28, 28, 3))
#     max_prob = max(result[0])
#     class_ind = list(result[0]).index(max_prob)
#     class_name = classes[class_ind]
#     print(class_name)
#     count+=1
#     if count>10:
#         break