def on_epoch_begin(self, epoch, logs=None):
for i in range(3):
plt.clf()
radmon_int = random.randint(0, test_dataset_2.__len__() - 1)
image, _ = test_dataset_2.__getitem__(radmon_int)
# print(image[0].shape)
same_l1 = detect_image(image[0], image[1], model=pred_model)
diff_l2 = detect_image(image[0], image[2], model=pred_model)
plt.subplot(1, 3, 1)
plt.imshow(np.array(image[0]))
plt.subplot(1, 3, 2)
plt.imshow(np.array(image[1]))
plt.text(-12,
-12,
'same:%.3f' % same_l1,
ha='center',
va='bottom',
fontsize=11)
plt.subplot(1, 3, 3)
plt.imshow(np.array(image[2]))
plt.text(-24,
-12,
'diff:%.3f' % diff_l2,
ha='center',
va='bottom',
fontsize=11)
plt.savefig(r'image\test_epoch_%s_%s.png' % (epoch, i))
def hello():
K.clear_session()
if request.method == 'POST':
algo = request.form['algo']
if 'file_input' not in request.files:
flash('No file part')
return redirect(request.url)
file = request.files['file_input']
if file.filename == '':
flash('No selected file')
return redirect(request.url)
if file and allowed_file(file.filename):
filename = secure_filename(file.filename)
# file.save(os.path.join(app.config['UPLOAD_FOLDER'], filename))
file_data = file.stream.read()
nparr = np.fromstring(file_data, np.uint8)
img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
if algo:
yolo = YOLO(0.6, 0.5, 'yolo.h5')
# result = detect_image(cv2.imread(os.path.join(app.config['UPLOAD_FOLDER'], filename)), yolo)
result = detect_image(img, yolo)
print(
result[1]
) # result will be list of tuple where each tuple is ( class, prob, [box coords])
img_str = cv2.imencode('.jpg', result[0])[1].tostring()
encoded = base64.b64encode(img_str).decode("utf-8")
mime = "image/jpg;"
out_image = f"data:{mime}base64,{encoded}"
# cv2.imwrite(os.path.join(app.config['RESULT_FOLDER'],filename), result[0])
else:
yolo = YOLO(0.8, 0.5, 'tiny_yolo.h5')
# result = detect_image(cv2.imread(os.path.join(app.config['UPLOAD_FOLDER'], filename)), yolo)
result = detect_image(img, yolo)
print(
result[1]
) # result will be list of tuple where each tuple is ( class, prob, [box coords])
img_str = cv2.imencode('.jpg', result[0])[1].tostring()
encoded = base64.b64encode(img_str).decode("utf-8")
mime = "image/jpg;"
out_image = f"data:{mime}base64,{encoded}"
# cv2.imwrite(os.path.join(app.config['RESULT_FOLDER'],filename), result[0])
return render_template('result.html', out_image=out_image)
else:
return "File extension not supported"
return render_template('index.html')
def post(self):
try:
model = load_model()
except:
print("Model Failed to load")
try:
nparr = np.fromstring(req.data, np.uint8)
img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
except:
print("Image failed to load")
try:
response = detect_image(img, model)
print(response[1:])
response_pickled = jsonpickle.encode(response)
except:
print("meh")
return Response(response=response_pickled,
status=200,
mimetype="application/json")
def detect_traffic_lights(PATH_TO_TEST_IMAGES_DIR, Num_images, plot_flag=False):
"""
Detect traffic lights and draw bounding boxes around the traffic lights
:param PATH_TO_TEST_IMAGES_DIR: testing image directory
:param MODEL_NAME: name of the model used in the task
:return: commands: True: go, False: stop
"""
# file_data = file.stream.read()
# nparr = np.fromstring(file_data, np.uint8)
# img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
TEST_IMAGE_PATHS = [os.path.join(PATH_TO_TEST_IMAGES_DIR, '{}'.format(
i)) for i in os.listdir(PATH_TO_TEST_IMAGES_DIR)]
for image_path in TEST_IMAGE_PATHS:
Image.MAX_IMAGE_PIXELS = None
image = Image.open(image_path)
image_np = load_image_into_numpy_array(image)
yolo = YOLO(0.6, 0.5)
# result = detect_image(cv2.imread(os.path.join(app.config['UPLOAD_FOLDER'], filename)), yolo)
result = detect_image(cv2.imread(image_path), yolo, coco_classes)
# result will be list of tuple where each tuple is ( class, score, [box coords])
classes = result[0]
scores = result[1]
boxes = result[2]
# print("Result :",result)
# print("Classes :",classes)
# print("Boxes :",boxes)
# print("Scores :",scores)
red_flag, crop_img = read_traffic_lights(image, np.array(boxes), np.array(scores), np.array(classes).astype(np.int32))
cv2.imwrite('images/res/' + image_path.rsplit('/', 1)
[-1], crop_img[..., ::-1])
if red_flag:
print('{}: stop'.format(image_path)) # red or yellow
else:
print('{}: go'.format(image_path))
import os os.environ['CUDA_VISIBLE_DEVICES'] = '0' import cv2 import numpy as np import tensorflow as tf from utils import detect_image, detect_realtime, detect_video, Load_Yolo_model, detect_video_realtime_mp from configs import * image_path = "" video_path = "" yolo = Load_Yolo_model() detect_image(yolo, image_path, "", input_size=YOLO_INPUT_SIZE, show=True, CLASSES=TRAIN_CLASSES, rectangle_colors=(255,0,0)) #detect_video(yolo, video_path, "", input_size=YOLO_INPUT_SIZE, show=False, CLASSES=TRAIN_CLASSES, rectangle_colors=(255,0,0)) #detect_realtime(yolo, "", input_size=YOLO_INPUT_SIZE, show=True, CLASSES=TRAIN_CLASSES, rectangle_colors=(255, 0, 0))
def update_dashboard(generate_dashboard):
#Download all the images from blob storage and processing with deeplearning model
images_div = []
blob_service_client = BlobServiceClient.from_connection_string(connect_str)
container_client = blob_service_client.get_container_client("pruebasds4")
my_blobs = container_client.list_blobs()
#Column name list
column_names = ['id', 'x1', 'y1', 'x2', 'y2', 'Score', 'Class']
#row list
rows = []
names = []
images = []
for blob in my_blobs:
image_downloaded = container_client.download_blob(blob.name).readall()
decoded_data = base64.b64decode(image_downloaded)
np_data = np.fromstring(decoded_data, np.uint8)
img = cv2.imdecode(np_data, cv2.IMREAD_UNCHANGED)
#INGRESARLAS AL MODELO
prediction = detect_image(
yolo,
img,
"./IMAGES/HU8_detect.jpg",
input_size=416,
show=True,
CLASSES="./model_data/license_plate_names.txt",
rectangle_colors=(255, 0, 0))
if prediction[1][0]:
for box_stats in prediction[1]:
box_stats['id'] = blob.name
rows.append(box_stats)
#CAST TO BASE64
retval, buffer_img = cv2.imencode('.jpg', prediction[0])
data = base64.b64encode(buffer_img)
images_div.append('data:image/png;base64,{}'.format(data.decode()))
#Create empty dataframe
df = pd.DataFrame(rows, columns=column_names)
df['Class'] = df['Class'].apply(lambda x: change_classname(x))
children = [generate_thumbnail(data) for data in images_div]
df2 = df.groupby(['Class']).size().reset_index(name='counts')
#GENERATE PLOTS
colors = ['#61892F', '#6B6E70', '#86C232']
ploty1 = go.Figure(
data=[go.Pie(labels=df2['Class'], values=df2['counts'], hole=.5)])
ploty1.update_layout(height=350,
width=470,
paper_bgcolor='rgba(0,0,0,0)',
plot_bgcolor='rgba(0,0,0,0)',
font=dict(family="Enriqueta, Times New Roman",
size=16,
color='rgb(97,137,47)'),
title="Faillure Distribution",
title_x=0.5)
ploty1.update_traces(marker=dict(colors=colors))
table = generate_table(df)
return children, table, ploty1
input_size = YOLO_INPUT_SIZE
path_in = os.path.join(PATH_DIR, "Images", "401-2")
save_to = os.path.join(PATH_DIR, "Images", "crop.jpg")
# Load Keras weights
yolo = Create_Yolov3(input_size=input_size, CLASSES=TRAIN_CLASSES)
yolo.load_weights("weights/checkpoints2/yolov3_custom")
# Detect bounding boxes in image
for i, filename in enumerate(glob.glob(path_in + r"\*.jpg")):
print(filename)
orig, img, locs = detect_image(yolo,
filename,
save_to,
input_size=input_size,
show=True,
CLASSES=TRAIN_CLASSES,
rectangle_colors=(255, 0, 0))
for j, crops in enumerate(locs):
new_img = orig[crops[1]:crops[3], crops[0]:crops[2], :3]
new_img = Image.fromarray(new_img)
w_size = int(
float(new_img.size[0]) * float(32 / float(new_img.size[1])))
new_img = new_img.resize((w_size, 32), Image.ANTIALIAS)
# path_ext = str(i+1200) + "-" + str(j) + ".jpg"
# new_img.save(os.path.join(r"C:\Users\sid_a\PycharmProjects\MTO_Highway_Analysis\time_stamps", path_ext))
print(i)
import matplotlib.pyplot as plt
import numpy as np
os.environ["CUDA_VISIBLE_DEVICES"] = "-1"
test_image_path = r'E:\DataSets\DukeMTMC-reID\DukeMTMC-reID\query'
test_dataset_2 = Person_Dataset(test_image_path, batch_size=1, train=False)
input_size = (215, 90, 3)
model, pred_model = Create_Model(inpt=input_size, num_classes=1812)
model.load_weights('logs\ep039-loss0.066.h5')
true_ = 0
for i in range(test_dataset_2.__len__()):
image, _ = test_dataset_2.__getitem__(i)
t1 = time.time()
same_l1 = detect_image(image[0], image[1], model=pred_model)
t2 = time.time()
diff_l2 = detect_image(image[0], image[2], model=pred_model)
plt.subplot(1, 3, 1)
plt.imshow(np.array(image[0]))
plt.subplot(1, 3, 2)
plt.imshow(np.array(image[1]))
plt.text(-12,
-12,
'same:%.3f' % same_l1,
ha='center',
va='bottom',
fontsize=11)