mirror of
https://github.com/blakeblackshear/frigate.git
synced 2024-11-21 19:07:46 +01:00
basic motion detection working
This commit is contained in:
parent
d8f27658f8
commit
5d894f006a
@ -1,5 +1,6 @@
|
|||||||
import os
|
import os
|
||||||
import cv2
|
import cv2
|
||||||
|
import imutils
|
||||||
import time
|
import time
|
||||||
import datetime
|
import datetime
|
||||||
import ctypes
|
import ctypes
|
||||||
@ -24,8 +25,8 @@ PATH_TO_LABELS = '/label_map.pbtext'
|
|||||||
# TODO: make dynamic?
|
# TODO: make dynamic?
|
||||||
NUM_CLASSES = 90
|
NUM_CLASSES = 90
|
||||||
|
|
||||||
#REGIONS = "600,0,380:600,600,380:600,1200,380"
|
REGIONS = "300,0,0:300,300,0:300,600,0"
|
||||||
REGIONS = os.getenv('REGIONS')
|
#REGIONS = os.getenv('REGIONS')
|
||||||
|
|
||||||
DETECTED_OBJECTS = []
|
DETECTED_OBJECTS = []
|
||||||
|
|
||||||
@ -152,62 +153,77 @@ def main():
|
|||||||
detection_process.daemon = True
|
detection_process.daemon = True
|
||||||
detection_processes.append(detection_process)
|
detection_processes.append(detection_process)
|
||||||
|
|
||||||
|
motion_processes = []
|
||||||
|
for index, region in enumerate(regions):
|
||||||
|
motion_process = mp.Process(target=detect_motion, args=(shared_arr,
|
||||||
|
shared_memory_objects[index]['frame_time'],
|
||||||
|
shared_memory_objects[index]['motion_detected'],
|
||||||
|
frame_shape,
|
||||||
|
region['size'], region['x_offset'], region['y_offset']))
|
||||||
|
motion_process.daemon = True
|
||||||
|
motion_processes.append(motion_process)
|
||||||
|
|
||||||
object_parser = ObjectParser([obj['output_array'] for obj in shared_memory_objects])
|
object_parser = ObjectParser([obj['output_array'] for obj in shared_memory_objects])
|
||||||
object_parser.start()
|
# object_parser.start()
|
||||||
|
|
||||||
capture_process.start()
|
capture_process.start()
|
||||||
print("capture_process pid ", capture_process.pid)
|
print("capture_process pid ", capture_process.pid)
|
||||||
for detection_process in detection_processes:
|
# for detection_process in detection_processes:
|
||||||
detection_process.start()
|
# detection_process.start()
|
||||||
print("detection_process pid ", detection_process.pid)
|
# print("detection_process pid ", detection_process.pid)
|
||||||
|
for motion_process in motion_processes:
|
||||||
|
motion_process.start()
|
||||||
|
print("motion_process pid ", motion_process.pid)
|
||||||
|
|
||||||
app = Flask(__name__)
|
# app = Flask(__name__)
|
||||||
|
|
||||||
@app.route('/')
|
# @app.route('/')
|
||||||
def index():
|
# def index():
|
||||||
# return a multipart response
|
# # return a multipart response
|
||||||
return Response(imagestream(),
|
# return Response(imagestream(),
|
||||||
mimetype='multipart/x-mixed-replace; boundary=frame')
|
# mimetype='multipart/x-mixed-replace; boundary=frame')
|
||||||
def imagestream():
|
# def imagestream():
|
||||||
global DETECTED_OBJECTS
|
# global DETECTED_OBJECTS
|
||||||
while True:
|
# while True:
|
||||||
# max out at 5 FPS
|
# # max out at 5 FPS
|
||||||
time.sleep(0.2)
|
# time.sleep(0.2)
|
||||||
# make a copy of the current detected objects
|
# # make a copy of the current detected objects
|
||||||
detected_objects = DETECTED_OBJECTS.copy()
|
# detected_objects = DETECTED_OBJECTS.copy()
|
||||||
# make a copy of the current frame
|
# # make a copy of the current frame
|
||||||
frame = frame_arr.copy()
|
# frame = frame_arr.copy()
|
||||||
# convert to RGB for drawing
|
# # convert to RGB for drawing
|
||||||
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
# frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
||||||
# draw the bounding boxes on the screen
|
# # draw the bounding boxes on the screen
|
||||||
for obj in DETECTED_OBJECTS:
|
# for obj in DETECTED_OBJECTS:
|
||||||
vis_util.draw_bounding_box_on_image_array(frame,
|
# vis_util.draw_bounding_box_on_image_array(frame,
|
||||||
obj['ymin'],
|
# obj['ymin'],
|
||||||
obj['xmin'],
|
# obj['xmin'],
|
||||||
obj['ymax'],
|
# obj['ymax'],
|
||||||
obj['xmax'],
|
# obj['xmax'],
|
||||||
color='red',
|
# color='red',
|
||||||
thickness=2,
|
# thickness=2,
|
||||||
display_str_list=["{}: {}%".format(obj['name'],int(obj['score']*100))],
|
# display_str_list=["{}: {}%".format(obj['name'],int(obj['score']*100))],
|
||||||
use_normalized_coordinates=False)
|
# use_normalized_coordinates=False)
|
||||||
|
|
||||||
for region in regions:
|
# for region in regions:
|
||||||
cv2.rectangle(frame, (region['x_offset'], region['y_offset']),
|
# cv2.rectangle(frame, (region['x_offset'], region['y_offset']),
|
||||||
(region['x_offset']+region['size'], region['y_offset']+region['size']),
|
# (region['x_offset']+region['size'], region['y_offset']+region['size']),
|
||||||
(255,255,255), 2)
|
# (255,255,255), 2)
|
||||||
# convert back to BGR
|
# # convert back to BGR
|
||||||
frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
|
# frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
|
||||||
# encode the image into a jpg
|
# # encode the image into a jpg
|
||||||
ret, jpg = cv2.imencode('.jpg', frame)
|
# ret, jpg = cv2.imencode('.jpg', frame)
|
||||||
yield (b'--frame\r\n'
|
# yield (b'--frame\r\n'
|
||||||
b'Content-Type: image/jpeg\r\n\r\n' + jpg.tobytes() + b'\r\n\r\n')
|
# b'Content-Type: image/jpeg\r\n\r\n' + jpg.tobytes() + b'\r\n\r\n')
|
||||||
|
|
||||||
app.run(host='0.0.0.0', debug=False)
|
# app.run(host='0.0.0.0', debug=False)
|
||||||
|
|
||||||
capture_process.join()
|
capture_process.join()
|
||||||
for detection_process in detection_processes:
|
# for detection_process in detection_processes:
|
||||||
detection_process.join()
|
# detection_process.join()
|
||||||
object_parser.join()
|
for motion_process in motion_processes:
|
||||||
|
motion_process.join()
|
||||||
|
# object_parser.join()
|
||||||
|
|
||||||
# convert shared memory array into numpy array
|
# convert shared memory array into numpy array
|
||||||
def tonumpyarray(mp_arr):
|
def tonumpyarray(mp_arr):
|
||||||
@ -307,6 +323,91 @@ def process_frames(shared_arr, shared_output_arr, shared_frame_time, shared_moti
|
|||||||
# copy the detected objects to the output array, filling the array when needed
|
# copy the detected objects to the output array, filling the array when needed
|
||||||
shared_output_arr[:] = objects + [0.0] * (60-len(objects))
|
shared_output_arr[:] = objects + [0.0] * (60-len(objects))
|
||||||
|
|
||||||
|
# do the actual object detection
|
||||||
|
def detect_motion(shared_arr, shared_frame_time, shared_motion, frame_shape, region_size, region_x_offset, region_y_offset):
|
||||||
|
# shape shared input array into frame for processing
|
||||||
|
arr = tonumpyarray(shared_arr).reshape(frame_shape)
|
||||||
|
|
||||||
|
no_frames_available = -1
|
||||||
|
avg_frame = None
|
||||||
|
last_motion = -1
|
||||||
|
while True:
|
||||||
|
now = datetime.datetime.now().timestamp()
|
||||||
|
# if it has been 30 seconds since the last motion, clear the flag
|
||||||
|
if last_motion > 0 and (now - last_motion) > 30:
|
||||||
|
last_motion = -1
|
||||||
|
shared_motion.value = 0
|
||||||
|
print("motion cleared")
|
||||||
|
# if there isnt a frame ready for processing
|
||||||
|
if shared_frame_time.value == 0.0:
|
||||||
|
# save the first time there were no frames available
|
||||||
|
if no_frames_available == -1:
|
||||||
|
no_frames_available = now
|
||||||
|
# if there havent been any frames available in 30 seconds,
|
||||||
|
# sleep to avoid using so much cpu if the camera feed is down
|
||||||
|
if no_frames_available > 0 and (now - no_frames_available) > 30:
|
||||||
|
time.sleep(1)
|
||||||
|
print("sleeping because no frames have been available in a while")
|
||||||
|
else:
|
||||||
|
# rest a little bit to avoid maxing out the CPU
|
||||||
|
time.sleep(0.01)
|
||||||
|
continue
|
||||||
|
|
||||||
|
# we got a valid frame, so reset the timer
|
||||||
|
no_frames_available = -1
|
||||||
|
|
||||||
|
# if the frame is more than 0.5 second old, discard it
|
||||||
|
if (now - shared_frame_time.value) > 0.5:
|
||||||
|
# signal that we need a new frame
|
||||||
|
shared_frame_time.value = 0.0
|
||||||
|
# rest a little bit to avoid maxing out the CPU
|
||||||
|
time.sleep(0.01)
|
||||||
|
continue
|
||||||
|
|
||||||
|
# make a copy of the cropped frame
|
||||||
|
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
|
||||||
|
# signal that the frame has been used so a new one will be ready
|
||||||
|
shared_frame_time.value = 0.0
|
||||||
|
|
||||||
|
# convert to grayscale
|
||||||
|
gray = cv2.cvtColor(cropped_frame, cv2.COLOR_BGR2GRAY)
|
||||||
|
# convert to uint8
|
||||||
|
gray = (gray/256).astype('uint8')
|
||||||
|
# apply gaussian blur
|
||||||
|
gray = cv2.GaussianBlur(gray, (21, 21), 0)
|
||||||
|
|
||||||
|
if avg_frame is None:
|
||||||
|
avg_frame = gray.copy().astype("float")
|
||||||
|
continue
|
||||||
|
|
||||||
|
# look at the delta from the avg_frame
|
||||||
|
cv2.accumulateWeighted(gray, avg_frame, 0.5)
|
||||||
|
frameDelta = cv2.absdiff(gray, cv2.convertScaleAbs(avg_frame))
|
||||||
|
thresh = cv2.threshold(frameDelta, 25, 255, cv2.THRESH_BINARY)[1]
|
||||||
|
|
||||||
|
# dilate the thresholded image to fill in holes, then find contours
|
||||||
|
# on thresholded image
|
||||||
|
thresh = cv2.dilate(thresh, None, iterations=2)
|
||||||
|
cnts = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL,
|
||||||
|
cv2.CHAIN_APPROX_SIMPLE)
|
||||||
|
cnts = imutils.grab_contours(cnts)
|
||||||
|
|
||||||
|
# loop over the contours
|
||||||
|
for c in cnts:
|
||||||
|
# if the contour is too small, ignore it
|
||||||
|
if cv2.contourArea(c) < 50:
|
||||||
|
continue
|
||||||
|
|
||||||
|
print("motion_detected")
|
||||||
|
last_motion = now
|
||||||
|
shared_motion.value = 1
|
||||||
|
|
||||||
|
# compute the bounding box for the contour, draw it on the frame,
|
||||||
|
# and update the text
|
||||||
|
(x, y, w, h) = cv2.boundingRect(c)
|
||||||
|
cv2.rectangle(cropped_frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
|
||||||
|
cv2.imwrite("motion%d.png" % frame_time, cropped_frame)
|
||||||
if __name__ == '__main__':
|
if __name__ == '__main__':
|
||||||
mp.freeze_support()
|
mp.freeze_support()
|
||||||
main()
|
main()
|
Loading…
Reference in New Issue
Block a user