blakeblackshear.frigate/frigate/video.py

367 lines
14 KiB
Python
Raw Permalink Normal View History

2019-03-30 02:49:27 +01:00
import os
2019-02-26 03:27:02 +01:00
import time
import datetime
import cv2
import queue
import threading
2019-03-30 02:49:27 +01:00
import ctypes
import pyarrow.plasma as plasma
2019-03-30 02:49:27 +01:00
import multiprocessing as mp
import subprocess as sp
2019-05-10 13:19:39 +02:00
import numpy as np
2020-01-11 20:22:56 +01:00
import copy
2019-12-31 21:59:22 +01:00
import itertools
import json
import base64
from typing import Dict, List
from collections import defaultdict
from frigate.util import draw_box_with_label, area, calculate_region, clipped, intersection_over_union, intersection, EventsPerSecond, listen, FrameManager, PlasmaFrameManager
2020-02-16 04:07:54 +01:00
from frigate.objects import ObjectTracker
from frigate.edgetpu import RemoteObjectDetector
from frigate.motion import MotionDetector
2019-02-26 03:27:02 +01:00
2019-12-08 14:03:58 +01:00
def get_frame_shape(source):
ffprobe_cmd = " ".join([
'ffprobe',
'-v',
'panic',
'-show_error',
'-show_streams',
'-of',
'json',
'"'+source+'"'
])
print(ffprobe_cmd)
p = sp.Popen(ffprobe_cmd, stdout=sp.PIPE, shell=True)
(output, err) = p.communicate()
p_status = p.wait()
info = json.loads(output)
print(info)
video_info = [s for s in info['streams'] if s['codec_type'] == 'video'][0]
if video_info['height'] != 0 and video_info['width'] != 0:
return (video_info['height'], video_info['width'], 3)
# fallback to using opencv if ffprobe didnt succeed
video = cv2.VideoCapture(source)
ret, frame = video.read()
frame_shape = frame.shape
video.release()
return frame_shape
2019-03-30 02:49:27 +01:00
2019-12-08 14:03:58 +01:00
def get_ffmpeg_input(ffmpeg_input):
frigate_vars = {k: v for k, v in os.environ.items() if k.startswith('FRIGATE_')}
return ffmpeg_input.format(**frigate_vars)
2019-03-30 02:49:27 +01:00
def filtered(obj, objects_to_track, object_filters, mask=None):
2020-02-16 04:07:54 +01:00
object_name = obj[0]
if not object_name in objects_to_track:
return True
if object_name in object_filters:
obj_settings = object_filters[object_name]
# if the min area is larger than the
# detected object, don't add it to detected objects
if obj_settings.get('min_area',-1) > obj[3]:
return True
# if the detected object is larger than the
# max area, don't add it to detected objects
if obj_settings.get('max_area', 24000000) < obj[3]:
return True
2020-09-07 19:17:42 +02:00
# if the score is lower than the min_score, skip
if obj_settings.get('min_score', 0) > obj[1]:
2020-02-16 04:07:54 +01:00
return True
# compute the coordinates of the object and make sure
# the location isnt outside the bounds of the image (can happen from rounding)
y_location = min(int(obj[2][3]), len(mask)-1)
x_location = min(int((obj[2][2]-obj[2][0])/2.0)+obj[2][0], len(mask[0])-1)
# if the object is in a masked location, don't add it to detected objects
2020-09-11 00:37:58 +02:00
if (not mask is None) and (mask[y_location][x_location] == 0):
2020-02-16 04:07:54 +01:00
return True
2020-09-07 19:17:42 +02:00
return False
2020-02-16 04:07:54 +01:00
def create_tensor_input(frame, region):
cropped_frame = frame[region[1]:region[3], region[0]:region[2]]
2019-03-30 02:49:27 +01:00
2020-02-16 04:07:54 +01:00
# Resize to 300x300 if needed
if cropped_frame.shape != (300, 300, 3):
cropped_frame = cv2.resize(cropped_frame, dsize=(300, 300), interpolation=cv2.INTER_LINEAR)
# Expand dimensions since the model expects images to have shape: [1, 300, 300, 3]
return np.expand_dims(cropped_frame, axis=0)
def start_or_restart_ffmpeg(ffmpeg_cmd, frame_size, ffmpeg_process=None):
if not ffmpeg_process is None:
print("Terminating the existing ffmpeg process...")
ffmpeg_process.terminate()
try:
print("Waiting for ffmpeg to exit gracefully...")
ffmpeg_process.communicate(timeout=30)
except sp.TimeoutExpired:
print("FFmpeg didnt exit. Force killing...")
ffmpeg_process.kill()
ffmpeg_process.communicate()
2020-03-13 21:50:27 +01:00
ffmpeg_process = None
print("Creating ffmpeg process...")
print(" ".join(ffmpeg_cmd))
process = sp.Popen(ffmpeg_cmd, stdout = sp.PIPE, stdin = sp.DEVNULL, bufsize=frame_size*10, start_new_session=True)
return process
def capture_frames(ffmpeg_process, camera_name, frame_shape, frame_manager: FrameManager,
frame_queue, take_frame: int, fps:EventsPerSecond, skipped_fps: EventsPerSecond,
2020-09-07 19:17:42 +02:00
stop_event: mp.Event, detection_frame: mp.Value, current_frame: mp.Value):
frame_num = 0
last_frame = 0
frame_size = frame_shape[0] * frame_shape[1] * frame_shape[2]
skipped_fps.start()
while True:
if stop_event.is_set():
print(f"{camera_name}: stop event set. exiting capture thread...")
break
frame_bytes = ffmpeg_process.stdout.read(frame_size)
2020-09-07 19:17:42 +02:00
current_frame.value = datetime.datetime.now().timestamp()
2020-09-13 17:15:08 +02:00
if len(frame_bytes) < frame_size:
print(f"{camera_name}: ffmpeg sent a broken frame. something is wrong.")
if ffmpeg_process.poll() != None:
print(f"{camera_name}: ffmpeg process is not running. exiting capture thread...")
break
else:
continue
fps.update()
frame_num += 1
if (frame_num % take_frame) != 0:
skipped_fps.update()
continue
# if the detection process is more than 1 second behind, skip this frame
if detection_frame.value > 0.0 and (last_frame - detection_frame.value) > 1:
skipped_fps.update()
continue
# put the frame in the frame manager
2020-09-07 19:17:42 +02:00
frame_manager.put(f"{camera_name}{current_frame.value}",
np
.frombuffer(frame_bytes, np.uint8)
.reshape(frame_shape)
)
# add to the queue
2020-09-07 19:17:42 +02:00
frame_queue.put(current_frame.value)
last_frame = current_frame.value
class CameraCapture(threading.Thread):
2020-08-02 15:46:36 +02:00
def __init__(self, name, ffmpeg_process, frame_shape, frame_queue, take_frame, fps, detection_frame, stop_event):
threading.Thread.__init__(self)
self.name = name
self.frame_shape = frame_shape
self.frame_size = frame_shape[0] * frame_shape[1] * frame_shape[2]
self.frame_queue = frame_queue
self.take_frame = take_frame
self.fps = fps
self.skipped_fps = EventsPerSecond()
self.plasma_client = PlasmaFrameManager(stop_event)
self.ffmpeg_process = ffmpeg_process
2020-09-07 19:17:42 +02:00
self.current_frame = mp.Value('d', 0.0)
self.last_frame = 0
self.detection_frame = detection_frame
2020-08-02 15:46:36 +02:00
self.stop_event = stop_event
def run(self):
self.skipped_fps.start()
capture_frames(self.ffmpeg_process, self.name, self.frame_shape, self.plasma_client, self.frame_queue, self.take_frame,
2020-09-07 19:17:42 +02:00
self.fps, self.skipped_fps, self.stop_event, self.detection_frame, self.current_frame)
2020-09-07 19:17:42 +02:00
def track_camera(name, config, frame_queue, frame_shape, detection_queue, detected_objects_queue, fps, detection_fps, read_start, detection_frame, stop_event):
2020-02-16 04:07:54 +01:00
print(f"Starting process for {name}: {os.getpid()}")
listen()
2020-02-16 04:07:54 +01:00
detection_frame.value = 0.0
2020-02-16 04:07:54 +01:00
# Merge the tracked object config with the global config
2020-09-07 19:17:42 +02:00
camera_objects_config = config.get('objects', {})
objects_to_track = camera_objects_config.get('track', [])
object_filters = camera_objects_config.get('filters', {})
2020-02-16 04:07:54 +01:00
# load in the mask for object detection
if 'mask' in config:
if config['mask'].startswith('base64,'):
img = base64.b64decode(config['mask'][7:])
npimg = np.fromstring(img, dtype=np.uint8)
mask = cv2.imdecode(npimg, cv2.IMREAD_GRAYSCALE)
2020-09-13 15:23:41 +02:00
elif config['mask'].startswith('poly,'):
points = config['mask'].split(',')[1:]
contour = np.array([[int(points[i]), int(points[i+1])] for i in range(0, len(points), 2)])
mask = np.zeros((frame_shape[0], frame_shape[1]), np.uint8)
mask[:] = 255
cv2.fillPoly(mask, pts=[contour], color=(0))
else:
mask = cv2.imread("/config/{}".format(config['mask']), cv2.IMREAD_GRAYSCALE)
2020-02-16 04:07:54 +01:00
else:
mask = None
2020-09-13 05:56:19 +02:00
if mask is None or mask.size == 0:
2020-09-11 00:37:58 +02:00
mask = np.zeros((frame_shape[0], frame_shape[1]), np.uint8)
2020-02-16 04:07:54 +01:00
mask[:] = 255
motion_detector = MotionDetector(frame_shape, mask, resize_factor=6)
object_detector = RemoteObjectDetector(name, '/labelmap.txt', detection_queue)
2020-02-16 04:07:54 +01:00
object_tracker = ObjectTracker(10)
plasma_client = PlasmaFrameManager()
process_frames(name, frame_queue, frame_shape, plasma_client, motion_detector, object_detector,
2020-09-13 05:29:53 +02:00
object_tracker, detected_objects_queue, fps, detection_fps, detection_frame, objects_to_track, object_filters, mask, stop_event)
print(f"{name}: exiting subprocess")
def reduce_boxes(boxes):
if len(boxes) == 0:
return []
reduced_boxes = cv2.groupRectangles([list(b) for b in itertools.chain(boxes, boxes)], 1, 0.2)[0]
return [tuple(b) for b in reduced_boxes]
def detect(object_detector, frame, region, objects_to_track, object_filters, mask):
tensor_input = create_tensor_input(frame, region)
detections = []
region_detections = object_detector.detect(tensor_input)
for d in region_detections:
box = d[2]
size = region[2]-region[0]
x_min = int((box[1] * size) + region[0])
y_min = int((box[0] * size) + region[1])
x_max = int((box[3] * size) + region[0])
y_max = int((box[2] * size) + region[1])
det = (d[0],
d[1],
(x_min, y_min, x_max, y_max),
(x_max-x_min)*(y_max-y_min),
region)
# apply object filters
if filtered(det, objects_to_track, object_filters, mask):
continue
detections.append(det)
return detections
def process_frames(camera_name: str, frame_queue: mp.Queue, frame_shape,
frame_manager: FrameManager, motion_detector: MotionDetector,
object_detector: RemoteObjectDetector, object_tracker: ObjectTracker,
2020-09-13 05:29:53 +02:00
detected_objects_queue: mp.Queue, fps: mp.Value, detection_fps: mp.Value, current_frame_time: mp.Value,
objects_to_track: List[str], object_filters: Dict, mask, stop_event: mp.Event,
exit_on_empty: bool = False):
2020-02-16 04:07:54 +01:00
fps_tracker = EventsPerSecond()
fps_tracker.start()
2020-02-16 04:07:54 +01:00
while True:
if stop_event.is_set() or (exit_on_empty and frame_queue.empty()):
print(f"Exiting track_objects...")
break
try:
frame_time = frame_queue.get(True, 10)
except queue.Empty:
continue
current_frame_time.value = frame_time
frame = frame_manager.get(f"{camera_name}{frame_time}")
if frame is None:
print(f"{camera_name}: frame {frame_time} is not in memory store.")
continue
fps_tracker.update()
fps.value = fps_tracker.eps()
2020-02-16 04:07:54 +01:00
# look for motion
motion_boxes = motion_detector.detect(frame)
tracked_object_boxes = [obj['box'] for obj in object_tracker.tracked_objects.values()]
# combine motion boxes with known locations of existing objects
combined_boxes = reduce_boxes(motion_boxes + tracked_object_boxes)
# compute regions
regions = [calculate_region(frame_shape, a[0], a[1], a[2], a[3], 1.2)
for a in combined_boxes]
# combine overlapping regions
combined_regions = reduce_boxes(regions)
2020-02-16 04:07:54 +01:00
# re-compute regions
regions = [calculate_region(frame_shape, a[0], a[1], a[2], a[3], 1.0)
for a in combined_regions]
2020-02-16 04:07:54 +01:00
# resize regions and detect
detections = []
for region in regions:
detections.extend(detect(object_detector, frame, region, objects_to_track, object_filters, mask))
2020-02-16 04:07:54 +01:00
#########
# merge objects, check for clipped objects and look again up to 4 times
2020-02-16 04:07:54 +01:00
#########
refining = True
refine_count = 0
while refining and refine_count < 4:
refining = False
# group by name
detected_object_groups = defaultdict(lambda: [])
for detection in detections:
detected_object_groups[detection[0]].append(detection)
selected_objects = []
for group in detected_object_groups.values():
# apply non-maxima suppression to suppress weak, overlapping bounding boxes
boxes = [(o[2][0], o[2][1], o[2][2]-o[2][0], o[2][3]-o[2][1])
for o in group]
confidences = [o[1] for o in group]
idxs = cv2.dnn.NMSBoxes(boxes, confidences, 0.5, 0.4)
for index in idxs:
obj = group[index[0]]
if clipped(obj, frame_shape):
2020-02-16 04:07:54 +01:00
box = obj[2]
# calculate a new region that will hopefully get the entire object
region = calculate_region(frame_shape,
box[0], box[1],
box[2], box[3])
selected_objects.extend(detect(object_detector, frame, region, objects_to_track, object_filters, mask))
2020-02-16 04:07:54 +01:00
refining = True
else:
selected_objects.append(obj)
2020-02-16 04:07:54 +01:00
# set the detections list to only include top, complete objects
# and new detections
detections = selected_objects
if refining:
refine_count += 1
2020-02-16 04:07:54 +01:00
# now that we have refined our detections, we need to track objects
object_tracker.match_and_update(frame_time, detections)
# add to the queue
detected_objects_queue.put((camera_name, frame_time, object_tracker.tracked_objects))
2020-09-13 05:29:53 +02:00
detection_fps.value = object_detector.fps.eps()