Mirrors/blakeblackshear.frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2025-10-27 10:52:11 +01:00
Code Issues Projects Releases Wiki Activity

switch to a thread for object detection

Browse Source
This commit is contained in:
blakeblackshear 2019-03-27 20:44:57 -05:00
parent a074945394
commit 0514eeac03
2 changed files with 29 additions and 109 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
detect_objects.py
View File

@ -75,22 +75,12 @@ def main():
frame_lock = mp.Lock() frame_lock = mp.Lock()
# Condition for notifying that a new frame is ready # Condition for notifying that a new frame is ready
frame_ready = mp.Condition() frame_ready = mp.Condition()
# Shared memory array for passing prepped frame to tensorflow
prepped_frame_array = mp.Array(ctypes.c_uint8, 300*300*3)
# create shared value for storing the frame_time
prepped_frame_time = mp.Value('d', 0.0)
# Event for notifying that object detection needs a new frame
prepped_frame_grabbed = mp.Event()
# Event for notifying that new frame is ready for detection
prepped_frame_ready = mp.Event()
# Condition for notifying that objects were parsed # Condition for notifying that objects were parsed
objects_parsed = mp.Condition() objects_parsed = mp.Condition()
# Queue for detected objects # Queue for detected objects
object_queue = mp.Queue() object_queue = queue.Queue()
# Queue for prepped frames # Queue for prepped frames
prepped_frame_queue = queue.Queue(len(regions)*2) prepped_frame_queue = queue.Queue(len(regions)*2)
# Array for passing original region box to compute object bounding box
prepped_frame_box = mp.Array(ctypes.c_uint16, 3)
# shape current frame so it can be treated as an image # shape current frame so it can be treated as an image
frame_arr = tonumpyarray(shared_arr).reshape(frame_shape) frame_arr = tonumpyarray(shared_arr).reshape(frame_shape)
@ -113,28 +103,11 @@ def main():
)) ))
prepped_queue_processor = PreppedQueueProcessor( prepped_queue_processor = PreppedQueueProcessor(
prepped_frame_array, prepped_frame_queue,
prepped_frame_time, object_queue
prepped_frame_ready,
prepped_frame_grabbed,
prepped_frame_box,
prepped_frame_queue
) )
prepped_queue_processor.start() prepped_queue_processor.start()
# create a process for object detection
# if the coprocessor is doing the work, can this run as a thread
# since it is waiting for IO?
detection_process = mp.Process(target=detect_objects, args=(
prepped_frame_array,
prepped_frame_time,
prepped_frame_ready,
prepped_frame_grabbed,
prepped_frame_box,
object_queue, DEBUG
))
detection_process.daemon = True
# start a thread to store recent motion frames for processing # start a thread to store recent motion frames for processing
frame_tracker = FrameTracker(frame_arr, shared_frame_time, frame_ready, frame_lock, frame_tracker = FrameTracker(frame_arr, shared_frame_time, frame_ready, frame_lock,
recent_frames) recent_frames)
@ -177,9 +150,6 @@ def main():
for detection_prep_thread in detection_prep_threads: for detection_prep_thread in detection_prep_threads:
detection_prep_thread.start() detection_prep_thread.start()
detection_process.start()
print("detection_process pid ", detection_process.pid)
# 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__)
@ -237,7 +207,6 @@ def main():
capture_process.join() capture_process.join()
for detection_prep_thread in detection_prep_threads: for detection_prep_thread in detection_prep_threads:
detection_prep_thread.join() detection_prep_thread.join()
detection_process.join()
frame_tracker.join() frame_tracker.join()
best_person_frame.join() best_person_frame.join()
object_parser.join() object_parser.join()

93
frigate/object_detection.py
View File

@ -21,30 +21,24 @@ def ReadLabelFile(file_path):
ret[int(pair[0])] = pair[1].strip() ret[int(pair[0])] = pair[1].strip()
return ret return ret
def detect_objects(prepped_frame_array, prepped_frame_time, class PreppedQueueProcessor(threading.Thread):
prepped_frame_ready, prepped_frame_grabbed, def __init__(self, prepped_frame_queue, object_queue):
prepped_frame_box, object_queue, debug):
prepped_frame_np = tonumpyarray(prepped_frame_array) threading.Thread.__init__(self)
self.prepped_frame_queue = prepped_frame_queue
self.object_queue = object_queue
# Load the edgetpu engine and labels # Load the edgetpu engine and labels
engine = DetectionEngine(PATH_TO_CKPT) self.engine = DetectionEngine(PATH_TO_CKPT)
labels = ReadLabelFile(PATH_TO_LABELS) self.labels = ReadLabelFile(PATH_TO_LABELS)
frame_time = 0.0 def run(self):
region_box = [0,0,0] # process queue...
while True: while True:
# wait until a frame is ready frame = self.prepped_frame_queue.get()
prepped_frame_ready.wait() # print(self.prepped_frame_queue.qsize())
prepped_frame_copy = prepped_frame_np.copy()
frame_time = prepped_frame_time.value
region_box[:] = prepped_frame_box
prepped_frame_grabbed.set()
# print("Grabbed " + str(region_box[1]) + "," + str(region_box[2]))
# Actual detection. # Actual detection.
objects = engine.DetectWithInputTensor(prepped_frame_copy, threshold=0.5, top_k=3) objects = self.engine.DetectWithInputTensor(frame['frame'], threshold=0.5, top_k=3)
# time.sleep(0.1) # time.sleep(0.1)
# objects = [] # objects = []
# print(engine.get_inference_time()) # print(engine.get_inference_time())
@ -52,58 +46,15 @@ def detect_objects(prepped_frame_array, prepped_frame_time,
if objects: if objects:
for obj in objects: for obj in objects:
box = obj.bounding_box.flatten().tolist() box = obj.bounding_box.flatten().tolist()
object_queue.put({ self.object_queue.put({
'frame_time': frame_time, 'frame_time': frame['frame_time'],
'name': str(labels[obj.label_id]), 'name': str(self.labels[obj.label_id]),
'score': float(obj.score), 'score': float(obj.score),
'xmin': int((box[0] * region_box[0]) + region_box[1]), 'xmin': int((box[0] * frame['region_size']) + frame['region_x_offset']),
'ymin': int((box[1] * region_box[0]) + region_box[2]), 'ymin': int((box[1] * frame['region_size']) + frame['region_y_offset']),
'xmax': int((box[2] * region_box[0]) + region_box[1]), 'xmax': int((box[2] * frame['region_size']) + frame['region_x_offset']),
'ymax': int((box[3] * region_box[0]) + region_box[2]) 'ymax': int((box[3] * frame['region_size']) + frame['region_y_offset'])
}) })
# else:
# object_queue.put({
# 'frame_time': frame_time,
# 'name': 'dummy',
# 'score': 0.99,
# 'xmin': int(0 + region_box[1]),
# 'ymin': int(0 + region_box[2]),
# 'xmax': int(10 + region_box[1]),
# 'ymax': int(10 + region_box[2])
# })
class PreppedQueueProcessor(threading.Thread):
def __init__(self, prepped_frame_array,
prepped_frame_time,
prepped_frame_ready,
prepped_frame_grabbed,
prepped_frame_box,
prepped_frame_queue):
threading.Thread.__init__(self)
self.prepped_frame_array = prepped_frame_array
self.prepped_frame_time = prepped_frame_time
self.prepped_frame_ready = prepped_frame_ready
self.prepped_frame_grabbed = prepped_frame_grabbed
self.prepped_frame_box = prepped_frame_box
self.prepped_frame_queue = prepped_frame_queue
def run(self):
prepped_frame_np = tonumpyarray(self.prepped_frame_array)
# process queue...
while True:
frame = self.prepped_frame_queue.get()
# print(self.prepped_frame_queue.qsize())
prepped_frame_np[:] = frame['frame']
self.prepped_frame_time.value = frame['frame_time']
self.prepped_frame_box[0] = frame['region_size']
self.prepped_frame_box[1] = frame['region_x_offset']
self.prepped_frame_box[2] = frame['region_y_offset']
# print("Passed " + str(frame['region_x_offset']) + "," + str(frame['region_x_offset']))
self.prepped_frame_ready.set()
self.prepped_frame_grabbed.wait()
self.prepped_frame_grabbed.clear()
self.prepped_frame_ready.clear()
# should this be a region class? # should this be a region class?
@ -156,5 +107,5 @@ class FramePrepper(threading.Thread):
'region_x_offset': self.region_x_offset, 'region_x_offset': self.region_x_offset,
'region_y_offset': self.region_y_offset 'region_y_offset': self.region_y_offset
}) })
# else: else:
# print("queue full. moving on") print("queue full. moving on")