def image():
try:
image_file = request.files['image'] # get the image
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# Get the target object class to be detected (alternative: request.headers['target_class'])
target_class = request.form.get('target_class')
# finally run the image through tensor flow object detection
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, target_class,
threshold)
json_output = json.dumps(objects,
indent=4,
sort_keys=True,
default=str)
response = Response(response=json_output,
status=200,
mimetype="application/json")
return response
except Exception as e:
return Response(response=json.dumps({'error': e},
indent=4,
sort_keys=True,
default=str),
status=500,
mimetype="application/json")
def image():
try:
image_file = request.files['image'].stream # get the image
#dobox=1 if request.args.get('withbox','no')=='yes' else 0
#print('dobox',dobox)
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# finally run the image through tensor flow object detection`
image_object = Image.open(image_file)
im = image_object.copy().convert('RGB')
objects = object_detection_api.get_objects(im, threshold)
#if dobox:
# img = object_detection_api.get_objects_image(im)
# import base64
# from io import BytesIO
# buffered = BytesIO()
# img.save(buffered, format="JPEG")
# img_str = base64.b64encode(buffered.getvalue())
# imgt="""<img src="data:image/jpeg;base64,{0}"/>""".format(img_str)
# return json.dumps(objects)+"<br><br><br>"+imgt
return objects
except Exception as e:
print('POST /image error: %e' % e)
return e
def test():
TEST_IMAGE_PATHS = './test_images/image.jpg'
image = Image.open(TEST_IMAGE_PATHS)
objects = object_detection_api.get_objects(image)
return objects
def test():
PATH_TO_TEST_IMAGES_DIR = 'object_detection/test_images' # cwh
TEST_IMAGE_PATHS = [os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 3)]
image = Image.open(TEST_IMAGE_PATHS[0])
objects = object_detection_api.get_objects(image)
return objects
def test():
PATH_TO_TEST_IMAGES_DIR = 'object_detection/test_images' # cwh
TEST_IMAGE_PATHS = [os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 3)]
image = Image.open(TEST_IMAGE_PATHS[0])
objects = object_detection_api.get_objects(image)
return objects
def getObject():
startTime = time.time()
if request.method == 'GET':
threshold = request.args.get('threshold', 0.5)
file_name = request.args.get('file_name')
objects = object_detection_api.get_objects(file_name, threshold)
print("-------getObject total time: %f" % (time.time() - startTime))
return objects
return 'method error'
def check():
startTime = time.time()
if request.method == 'POST':
image_file = request.files['file']
base_path = os.path.abspath(os.path.dirname(__file__))
upload_path = os.path.join(base_path, 'static/upload/')
file_name = upload_path + image_file.filename
image_file.save(file_name)
threshold = float(request.form.get('threshold', 0.5))
objects = object_detection_api.get_objects(file_name, threshold)
return objects
return 'method error'
def test():
threshold = request.form.get('threshold')
PATH_TO_TEST_IMAGES_DIR = 'object_detection/test_images' # cwh
TEST_IMAGE_PATHS = [
os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i))
for i in range(1, 2)
][0]
threshold = 0.7
img = cv2.imread(TEST_IMAGE_PATHS)
outputJson = object_detection_api.get_objects(img, threshold)
print(outputJson)
return outputJson
def image():
startTime = time.time()
if request.method == 'POST':
image_file = request.files['file']
base_path = os.path.abspath(os.path.dirname(__file__))
upload_path = os.path.join(base_path, 'static/upload/')
file_name = upload_path + image_file.filename
image_file.save(file_name)
threshold = request.form.get('threshold', 0.5)
objects = object_detection_api.get_objects(file_name, threshold)
print("-------total time: %f" % (time.time() - startTime))
return render_template('index.html',
json_data=objects,
img=image_file.filename)
return 'method error'
def test():
try:
CWD_PATH = os.getcwd()
PATH_TO_TEST_IMAGES_DIR = CWD_PATH + '/images'
TEST_IMAGE_PATHS = [
os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i))
for i in range(1, 3)
]
image = Image.open(TEST_IMAGE_PATHS[0])
objects = object_detection_api.get_objects(image)
print('objects :', objects)
return objects
except Exception as e:
print('POST /test error: %e', e)
return e
def image():
try:
image_file = request.files['image']
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, threshold)
return objects
except Exception as e:
print('POST /image error: %e' % e)
return e
def image():
try:
image_file = request.files['image'].read() # get the image
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.7
else:
threshold = float(threshold)
img = np.array(Image.open(io.BytesIO(image_file)))
outputJson = object_detection_api.get_objects(img, threshold)
return outputJson
except Exception as e:
print('POST /image error: %e' % e)
return e
def image():
try:
image_file = request.files['image'] # get the image
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# finally run the image through tensor flow object detection`
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, threshold)
return objects
except Exception as e:
print('POST /image error: %e' % e)
return e
def image():
try:
image_file = request.files['image'] # get the image
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# finally run the image through tensor flow object detection`
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, threshold)
return objects
except Exception as e:
print('POST /image error: %e' % e)
return e
def process_image():
"""Process image."""
try:
# Get the image.
image_file = request.files['image']
# Set an image confidence threshold value to limit returned data.
threshold = request.form.get('threshold')
if threshold is None:
threshold = object_detection_api.THRESHOLD
else:
threshold = float(threshold)
# Run the image through Tensorflow object detection.
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, threshold)
return objects
except Exception as e:
print('POST /image error: %e' % e)
return e
def image():
try:
image_file = request.files['image'] # get the image
# print(image_file)
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
threshold = 0.6
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# finally run the image through tensor flow object detection`
image_object = Image.open(image_file)
image_object = np.array(image_object)
objects = object_detection_api.get_objects(image_object, threshold)
return objects
except Exception as e:
traceback.print_exc()
return 'An ERROR raised on image()'
return 'OK'
def image():
try:
image_file = request.files['image'] # get the image
# Set an image confidence threshold value to limit returned data
threshold = request.form.get('threshold')
if threshold is None:
threshold = 0.5
else:
threshold = float(threshold)
# finally run the image through tensor flow object detection`
image_object = Image.open(image_file)
objects = object_detection_api.get_objects(image_object, threshold)
return Response(response=objects,
status=200,
mimetype="application/json")
except Exception as e:
return Response(response=json.dumps({error: e},
indent=4,
sort_keys=True,
default=str),
status=500,
mimetype="application/json")
def test():
"""Test API."""
image = Image.open(TEST_IMAGE_PATHS[0])
objects = object_detection_api.get_objects(image)
return objects
import object_detection_api
import os
from PIL import Image
# If you want to test the code with your images, just add path to the images to the TEST_IMAGE_PATHS.
PATH_TO_TEST_IMAGES_DIR = 'object_detection/test_images' #cwh
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 3) ]
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
response = object_detection_api.get_objects(image)
print("returned JSON: \n%s" % response)
def detect_object(image_object, threshold):
return object_detection_api.get_objects(image_object, threshold)
def index():
base_path = os.path.abspath(os.path.dirname(__file__))
upload_path = os.path.join(base_path, 'static/')
file_name = upload_path + 'test.jpg'
objects = object_detection_api.get_objects(file_name)
return objects
import object_detection_api
import os
from PIL import Image
# If you want to test the code with your images, just add path to the images to the TEST_IMAGE_PATHS.
PATH_TO_TEST_IMAGES_DIR = 'C:/tensorflow1/models/research/object_detection/test_images' #cwh
TEST_IMAGE_PATHS = [
os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i))
for i in range(1, 3)
]
# for image_path in TEST_IMAGE_PATHS:
# image = Image.open(image_path)
# response = object_detection_api.get_objects(image)
# print("returned JSON: \n%s" % response)
image = Image.open(
'C:/tensorflow1/models/research/object_detection/test_images/image.jpg')
response = object_detection_api.get_objects(image)
print("returned JSON: \n%s" % response)
def webhook():
data = request.get_json(force=True)
app.logger.info(data)
update = telegram.update.Update.de_json(request.get_json(force=True), bot)
if update.message.text == '/help':
bot.sendMessage(chat_id=update.message.chat_id,
text=helpers.help_text())
return 'OK'
photo_id, status = helpers.get_is_photo(data)
app.logger.info(photo_id)
if status == 1:
bot.sendMessage(
chat_id=update.message.chat_id,
text=
'Sorry. Unsupported file type. Please use jpg (jpeg) or png file.')
return 'OK'
if photo_id:
photo = helpers.download_file(app, photo_id, 'photos/downloads/')
image = cv2.imread(photo)
orig_height, orig_width, _ = image.shape
objects = object_detection_api.get_objects(image, THRESHOLD)
objects = json.loads(objects)
if len(objects) > 1:
app.logger.info(objects)
font = cv2.FONT_HERSHEY_SIMPLEX
result_msg = ''
for item in objects:
if item['name'] != 'Object':
continue
x = int(orig_width * item['x'])
y = int(orig_height * item['y'])
width = int(orig_width * item['width'])
height = int(orig_height * item['height'])
cv2.rectangle(image, (x, y), (width, height), (0, 255, 0), 2)
scope = float('{:.2f}'.format(item['score'] * 100))
cv2.putText(image,
item['class_name'] + " - " + str(scope) + '%',
(x + 5, y + 20), font, 1, (255, 255, 255), 1,
cv2.LINE_AA)
result_msg += item['class_name'] + " - " + str(scope) + '% \n'
new_name = 'photos/detected/photo_detected_' + str(
photo_id) + '.jpg'
cv2.imwrite(new_name, image)
bot.sendPhoto(update.message.chat_id,
photo=open(new_name, 'rb'),
caption="Result")
bot.sendMessage(chat_id=update.message.chat_id, text=result_msg)
else:
bot.sendMessage(
chat_id=update.message.chat_id,
text='Sorry! No objects were found. Please try another photo.')
return 'OK'
bot.sendMessage(chat_id=update.message.chat_id,
text='Please send a photo for recognition!')
return 'OK'
def processing(app, chat_id, video, object_name, new_object_name, iter_count,
threshold):
chat_id = str(chat_id)
continue_count = 500
root_dir = 'data/' + chat_id + '/'
if video:
sys.path.append('./object_detection')
import object_detection_api
helpers.check_or_create_dirs(root_dir, new_object_name)
try:
cap = cv2.VideoCapture(video)
except Exception as e:
print(e)
sys.exit()
x = 0
iteration = 0
while cap.isOpened():
x += 1
if (x % continue_count) != 0:
continue
ret, frame = cap.read()
orig_height, orig_width, _ = frame.shape
objects = object_detection_api.get_objects(frame, threshold)
objects = json.loads(objects)
app.logger.info(objects)
if len(objects) > 1:
for item in objects:
if item['name'] != 'Object':
continue
if item['class_name'] != object_name:
continue
iteration += 1
file_name = str(int(time.time())) + '-data'
x = int(orig_width * item['x'])
y = int(orig_height * item['y'])
width = int(orig_width * item['width'])
height = int(orig_height * item['height'])
cv2.imwrite(
root_dir + 'images/' + new_object_name + '/' +
file_name + '.jpg', frame)
helpers.create_annotation(root_dir, new_object_name,
file_name, x, y, width, height,
orig_width, orig_height)
if iteration == iter_count:
break
zip_file = 'data/' + str(int(time.time())) + '_' + chat_id + '.zip'
zf = zipfile.ZipFile(zip_file, "w")
for dirname, subdirs, files in os.walk(root_dir):
if dirname == 'data':
continue
zf.write(dirname)
for filename in files:
zf.write(os.path.join(dirname, filename))
zf.close()
deleteFiles = []
deleteDirs = []
for root, dirs, files in os.walk(root_dir):
if root == 'data':
continue
for f in files:
deleteFiles.append(os.path.join(root, f))
for d in dirs:
deleteDirs.append(os.path.join(root, d))
for f in deleteFiles:
os.remove(f)
for d in deleteDirs:
if d == 'data':
continue
os.rmdir(d)
return True, zip_file
else:
return False, ''