Mirrors/blakeblackshear.frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2024-11-21 19:07:46 +01:00
Code Issues Projects Releases Wiki Activity

first working version, single region and motion detection disabled

Browse Source
This commit is contained in:
blakeblackshear 2019-03-17 09:03:52 -05:00
parent de9c3f4d74
commit 8bae05cfe2
6 changed files with 65 additions and 72 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
Dockerfile
View File

@ -26,20 +26,25 @@ RUN apt-get -qq update && apt-get -qq install --no-install-recommends -y python3
vim \ vim \
ffmpeg \ ffmpeg \
unzip \ unzip \
libusb-1.0-0-dev \
python3-setuptools \
python3-numpy \
zlib1g-dev \
libgoogle-glog-dev \
swig \
libunwind-dev \
libc++-dev \
libc++abi-dev \
build-essential \
&& rm -rf /var/lib/apt/lists/* && rm -rf /var/lib/apt/lists/*
# Install core packages # Install core packages
RUN wget -q -O /tmp/get-pip.py --no-check-certificate https://bootstrap.pypa.io/get-pip.py && python3 /tmp/get-pip.py RUN wget -q -O /tmp/get-pip.py --no-check-certificate https://bootstrap.pypa.io/get-pip.py && python3 /tmp/get-pip.py
RUN pip install -U pip \ RUN pip install -U pip \
numpy \ numpy \
pillow \
matplotlib \ matplotlib \
notebook \ notebook \
jupyter \
pandas \
moviepy \
tensorflow \
keras \
autovizwidget \
Flask \ Flask \
imutils \ imutils \
paho-mqtt paho-mqtt
@ -59,9 +64,6 @@ RUN cd /usr/local/src/ \
&& ldconfig \ && ldconfig \
&& rm -rf /usr/local/src/protobuf-3.5.1/ && rm -rf /usr/local/src/protobuf-3.5.1/
# Add dataframe display widget
RUN jupyter nbextension enable --py --sys-prefix widgetsnbextension
# Download & build OpenCV # Download & build OpenCV
RUN wget -q -P /usr/local/src/ --no-check-certificate https://github.com/opencv/opencv/archive/4.0.1.zip RUN wget -q -P /usr/local/src/ --no-check-certificate https://github.com/opencv/opencv/archive/4.0.1.zip
RUN cd /usr/local/src/ \ RUN cd /usr/local/src/ \
@ -75,6 +77,16 @@ RUN cd /usr/local/src/ \
&& make install \ && make install \
&& rm -rf /usr/local/src/opencv-4.0.1 && rm -rf /usr/local/src/opencv-4.0.1
# Download and install EdgeTPU libraries
RUN wget -q -O edgetpu_api.tar.gz --no-check-certificate http://storage.googleapis.com/cloud-iot-edge-pretrained-models/edgetpu_api.tar.gz
RUN tar xzf edgetpu_api.tar.gz \
&& cd python-tflite-source \
&& cp -p libedgetpu/libedgetpu_arm32_throttled.so /lib/arm-linux-gnueabihf/libedgetpu.so \
&& cp edgetpu/swig/compiled_so/_edgetpu_cpp_wrapper_arm32.so edgetpu/swig/_edgetpu_cpp_wrapper.so \
&& cp edgetpu/swig/compiled_so/edgetpu_cpp_wrapper.py edgetpu/swig/ \
&& python3 setup.py develop --user
# Minimize image size # Minimize image size
RUN (apt-get autoremove -y; \ RUN (apt-get autoremove -y; \
apt-get autoclean -y) apt-get autoclean -y)
@ -87,4 +99,7 @@ WORKDIR /opt/frigate/
ADD frigate frigate/ ADD frigate frigate/
COPY detect_objects.py . COPY detect_objects.py .
CMD ["python3", "-u", "detect_objects.py"] CMD ["python3", "-u", "detect_objects.py"]
# WORKDIR /python-tflite-source/edgetpu/
# CMD ["python3", "-u", "demo/classify_image.py", "--model", "test_data/mobilenet_v2_1.0_224_inat_bird_quant_edgetpu.tflite", "--label", "test_data/inat_bird_labels.txt", "--image", "test_data/parrot.jpg"]

13
detect_objects.py
View File

@ -72,7 +72,7 @@ def main():
# compute the flattened array length from the array shape # compute the flattened array length from the array shape
flat_array_length = frame_shape[0] * frame_shape[1] * frame_shape[2] flat_array_length = frame_shape[0] * frame_shape[1] * frame_shape[2]
# create shared array for storing the full frame image data # create shared array for storing the full frame image data
shared_arr = mp.Array(ctypes.c_uint16, flat_array_length) shared_arr = mp.Array(ctypes.c_uint8, flat_array_length)
# create shared value for storing the frame_time # create shared value for storing the frame_time
shared_frame_time = mp.Value('d', 0.0) shared_frame_time = mp.Value('d', 0.0)
# Lock to control access to the frame # Lock to control access to the frame
@ -173,9 +173,14 @@ def main():
print("detection_process pid ", detection_process.pid) print("detection_process pid ", detection_process.pid)
# start the motion detection processes # start the motion detection processes
for motion_process in motion_processes: # for motion_process in motion_processes:
motion_process.start() # motion_process.start()
print("motion_process pid ", motion_process.pid) # print("motion_process pid ", motion_process.pid)
for region in regions:
region['motion_detected'].set()
with motion_changed:
motion_changed.notify_all()
# create a flask app that encodes frames a mjpeg on demand # create a flask app that encodes frames a mjpeg on demand
app = Flask(__name__) app = Flask(__name__)

2
frigate/motion.py
View File

@ -34,7 +34,7 @@ def detect_motion(shared_arr, shared_frame_time, frame_lock, frame_ready, motion
# lock and make a copy of the cropped frame # lock and make a copy of the cropped frame
with frame_lock: with frame_lock:
cropped_frame = arr[region_y_offset:region_y_offset+region_size, region_x_offset:region_x_offset+region_size].copy().astype('uint8') cropped_frame = arr[region_y_offset:region_y_offset+region_size, region_x_offset:region_x_offset+region_size].copy()
frame_time = shared_frame_time.value frame_time = shared_frame_time.value
# convert to grayscale # convert to grayscale

83
frigate/object_detection.py
View File

@ -1,9 +1,8 @@
import datetime import datetime
import cv2 import cv2
import numpy as np import numpy as np
import tensorflow as tf from edgetpu.detection.engine import DetectionEngine
from object_detection.utils import label_map_util from PIL import Image
from object_detection.utils import visualization_utils as vis_util
from . util import tonumpyarray from . util import tonumpyarray
# TODO: make dynamic? # TODO: make dynamic?
@ -13,58 +12,38 @@ PATH_TO_CKPT = '/frozen_inference_graph.pb'
# List of the strings that is used to add correct label for each box. # List of the strings that is used to add correct label for each box.
PATH_TO_LABELS = '/label_map.pbtext' PATH_TO_LABELS = '/label_map.pbtext'
# Loading label map # Function to read labels from text files.
label_map = label_map_util.load_labelmap(PATH_TO_LABELS) def ReadLabelFile(file_path):
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, with open(file_path, 'r') as f:
use_display_name=True) lines = f.readlines()
category_index = label_map_util.create_category_index(categories) ret = {}
for line in lines:
pair = line.strip().split(maxsplit=1)
ret[int(pair[0])] = pair[1].strip()
return ret
# do the actual object detection # do the actual object detection
def tf_detect_objects(cropped_frame, sess, detection_graph, region_size, region_x_offset, region_y_offset, debug): def tf_detect_objects(cropped_frame, engine, labels, region_size, region_x_offset, region_y_offset, debug):
# Resize to 300x300
cropped_frame = cv2.resize(cropped_frame, dsize=(300, 300), interpolation=cv2.INTER_LINEAR)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3] # Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(cropped_frame, axis=0) image_np_expanded = np.expand_dims(cropped_frame, axis=0)
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.
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.
scores = detection_graph.get_tensor_by_name('detection_scores:0')
classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
# Actual detection. # Actual detection.
(boxes, scores, classes, num_detections) = sess.run( ans = engine.DetectWithInputTensor(image_np_expanded.flatten(), threshold=0.5, top_k=3)
[boxes, scores, classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
if debug:
if len([value for index,value in enumerate(classes[0]) if str(category_index.get(value).get('name')) == 'person' and scores[0,index] > 0.5]) > 0:
vis_util.visualize_boxes_and_labels_on_image_array(
cropped_frame,
np.squeeze(boxes),
np.squeeze(classes).astype(np.int32),
np.squeeze(scores),
category_index,
use_normalized_coordinates=True,
line_thickness=4)
cv2.imwrite("/lab/debug/obj-{}-{}-{}.jpg".format(region_x_offset, region_y_offset, datetime.datetime.now().timestamp()), cropped_frame)
# build an array of detected objects # build an array of detected objects
objects = [] objects = []
for index, value in enumerate(classes[0]): if ans:
score = scores[0, index] for obj in ans:
if score > 0.5: box = obj.bounding_box.flatten().tolist()
box = boxes[0, index].tolist()
objects.append({ objects.append({
'name': str(category_index.get(value).get('name')), 'name': str(labels[obj.label_id]),
'score': float(score), 'score': float(obj.score),
'ymin': int((box[0] * region_size) + region_y_offset), 'xmin': int((box[0] * region_size) + region_x_offset),
'xmin': int((box[1] * region_size) + region_x_offset), 'ymin': int((box[1] * region_size) + region_y_offset),
'ymax': int((box[2] * region_size) + region_y_offset), 'xmax': int((box[2] * region_size) + region_x_offset),
'xmax': int((box[3] * region_size) + region_x_offset) 'ymax': int((box[3] * region_size) + region_y_offset)
}) })
return objects return objects
@ -75,15 +54,9 @@ def detect_objects(shared_arr, object_queue, shared_frame_time, frame_lock, fram
# shape shared input array into frame for processing # shape shared input array into frame for processing
arr = tonumpyarray(shared_arr).reshape(frame_shape) arr = tonumpyarray(shared_arr).reshape(frame_shape)
# Load a (frozen) Tensorflow model into memory before the processing loop # Load the edgetpu engine and labels
detection_graph = tf.Graph() engine = DetectionEngine(PATH_TO_CKPT)
with detection_graph.as_default(): labels = ReadLabelFile(PATH_TO_LABELS)
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
sess = tf.Session(graph=detection_graph)
frame_time = 0.0 frame_time = 0.0
while True: while True:
@ -105,7 +78,7 @@ def detect_objects(shared_arr, object_queue, shared_frame_time, frame_lock, fram
# convert to RGB # convert to RGB
cropped_frame_rgb = cv2.cvtColor(cropped_frame, cv2.COLOR_BGR2RGB) cropped_frame_rgb = cv2.cvtColor(cropped_frame, cv2.COLOR_BGR2RGB)
# do the object detection # do the object detection
objects = tf_detect_objects(cropped_frame_rgb, sess, detection_graph, region_size, region_x_offset, region_y_offset, debug) objects = tf_detect_objects(cropped_frame_rgb, engine, labels, region_size, region_x_offset, region_y_offset, debug)
for obj in objects: for obj in objects:
# ignore persons below the size threshold # ignore persons below the size threshold
if obj['name'] == 'person' and (obj['xmax']-obj['xmin'])*(obj['ymax']-obj['ymin']) < min_person_area: if obj['name'] == 'person' and (obj['xmax']-obj['xmin'])*(obj['ymax']-obj['ymin']) < min_person_area:

2
frigate/util.py
View File

@ -2,4 +2,4 @@ import numpy as np
# convert shared memory array into numpy array # convert shared memory array into numpy array
def tonumpyarray(mp_arr): def tonumpyarray(mp_arr):
return np.frombuffer(mp_arr.get_obj(), dtype=np.uint16) return np.frombuffer(mp_arr.get_obj(), dtype=np.uint8)

2
frigate/video.py
View File

@ -78,7 +78,7 @@ class FrameTracker(threading.Thread):
# lock and make a copy of the frame # lock and make a copy of the frame
with self.frame_lock: with self.frame_lock:
frame = self.shared_frame.copy().astype('uint8') frame = self.shared_frame.copy()
frame_time = self.frame_time.value frame_time = self.frame_time.value
# add the frame to recent frames # add the frame to recent frames