DR-App / app.py
pat229988's picture
Update app.py
f763a6b
import streamlit as st
from PIL import Image
import numpy as np
#import tensorflow as tf
from tensorflow import Graph as Graph
from tensorflow import import_graph_def
from tensorflow.compat.v1 import GraphDef as GraphDef
from tensorflow.compat.v1 import Session as Session
from tensorflow.io.gfile import GFile as GFile
from object_detection.utils import visualization_utils as vis_util
from object_detection.utils import label_map_util
# What model to download.
MODEL_NAME = 'E:\AIML-\Diabetic-Ratinopathy-master\optic_disc_macula_graph'
# Path to frozen detection graph. This is the actual model that is used for the object detection.
# PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'
PATH_TO_CKPT = 'resnet-inference-graph.pb'
NUM_CLASSES = 2
detection_graph = Graph()
with detection_graph.as_default():
od_graph_def = GraphDef()
with GFile(PATH_TO_CKPT, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
import_graph_def(od_graph_def, name='')
def load_image_into_numpy_array(image):
(im_width, im_height) = image.size
return np.array(image.getdata()).reshape(
(im_height, im_width, 3)).astype(np.uint8)
labelmap = {1: {'id': 1, 'name': 'optic_disease'}, 2: {'id': 2, 'name': 'macula'}}
dmp =[]
def pred(img):
with detection_graph.as_default():
with Session(graph=detection_graph) as sess:
# Definite input and output Tensors for detection_graph
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
# Each box represents a part of the image where a particular object was detected.
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
# Each score represent how level of confidence for each of the objects.
# Score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
# for image_path in img:
# image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(img)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
# Actual detection.
(boxes, scores, classes, num) = sess.run(
[detection_boxes, detection_scores, detection_classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
dmp.append([boxes, scores, classes, num])
# Visualization of the results of a detection.
vis_util.visualize_boxes_and_labels_on_image_array(
image_np,
np.squeeze(boxes),
np.squeeze(classes).astype(np.int32),
np.squeeze(scores),
# category_index,
labelmap,
use_normalized_coordinates=True,
line_thickness=40)
# plt.figure(figsize=(24,16))
# x = image_path.split("\\")
# x = list(map(lambda x: x.replace('tst_img', 'res_img'), x))
# fn = '//'.join(x)
# plt.imsave(fn,image_np)
# plt.imshow(image_np)
# plt.imsave(fn,image_np)
return(image_np,scores[0][0]*100,scores[0][1]*100)
#User Interface---------------------------------------------------------
uploaded_file = st.file_uploader("", type=['jpg','png','jpeg'])
pred_flag = False
def main():
st.label_visibility='collapse'
st.title("Diabetic Retinopathy Prediction")
if uploaded_file is not None:
image = Image.open(uploaded_file)
st.markdown('<p style="text-align: center;"><label>Image : </label></p>',unsafe_allow_html=True)
st.image(image,width=500)
if st.button("Predict"):
x,optic,macula = pred(image)
st.markdown('<p style="text-align: center;"><label>Prediction : </label></p>',unsafe_allow_html=True)
st.image(x,width=900)
st.write('Optic Disease Score',optic)
st.write('Macula Score',macula)
# result =''
# st.success('The output is {}'.format(result))
if __name__ == '__main__': #
main()