def main():
# Max frame number to detect and track
Max_frames=800
# video file name
video_path='TownCentreXVID.avi'
# label string list given by users.
labelstring=['good person', 'bad person','Good person']
# video capture
vidcap = cv2.VideoCapture(video_path)
#args = parse_args()
display = True
#use_dlibTracker = args.use_dlibTracker
saver = False
total_time = 0.0
total_frames = 0
# for disp
if display:
colours = np.random.rand(32, 3) # used only for display
# plt.ion()
# fig = plt.figure(figsize=(10,5))
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(use_dlib= False) #create instance of the kalman tracker
frames = detector.get_total_frames()
for frame in range(0, frames): # frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame)
# get the total frames
total_frames += 1
if (total_frames > Max_frames):
break
# get one fram from video
success, img = vidcap.read()
# fn = 'test/frame%d.jpg' % (frame) # video frames are extracted to 'test/Pictures%d.jpg' with ffmpeg
# img = io.imread(fn)
# if (display):
# ax1 = fig.add_subplot(111, aspect='equal')
# ax1.imshow(img)
start_time = time.time()
# update tracker
trackers = tracker.update(detections, img)
# time tracking
cycle_time = time.time() - start_time
total_time += cycle_time
print('frame: %d...took: %3fs' % (frame, cycle_time))
# object tracking
for d in trackers:
#f_out.write('%d,%d,%d,%d,x,x,x,x,%.3f,%.3f,%.3f,%.3f\n' % (d[4], frame, 1, 1, d[0], d[1], d[2], d[3]))
if (display):
d = d.astype(np.int32)
color, label_string = random_color(d[4], labelstring)
# add rectangle of object
cv2.rectangle(img, (d[0], d[1]), (d[2], d[3]), color, 3)
# ax1.add_patch(patches.Rectangle((d[0], d[1]), d[2] - d[0], d[3] - d[1], fill=False, lw=3,
# ec=color))
# ax1.set_adjustable('box-forced')
# add label string
font = cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(img, label_string, (d[0], d[1]), font, 1, color, 2, cv2.LINE_4)
# ax1.annotate(label_string, xy=(d[0], d[1]), xytext=(d[0], d[1]))
# if detections != []: # detector is active in this frame
# ax1.annotate(" DETECTOR", xy=(5, 45), xytext=(5, 45))
if (display):
# plt.axis('off')
# #fig.canvas.flush_events()
# plt.draw()
# fig.tight_layout()
cv2.imshow("frame",img)
cv2.waitKey(0)
# save the frame with tracking boxes
if (saver):
cv2.imwrite("frameout/f"+str(frame)+".jpg",img)
# fig.set_size_inches(18.5, 10.5)
# fig.savefig("frameout/f%d.jpg" % frame, dpi=200)
# ax1.cla()
print("Total Tracking took: %.3f for %d frames or %.1f FPS"%(total_time,total_frames,total_frames/total_time))
def main():
args = parse_args()
display = args.display
#display = False
use_dlibTracker = args.use_dlibTracker
#use_dlibTracker = False
saver = args.saver
total_time = 0.0
total_frames = 0
# for disp
if display:
colours = np.random.rand(32, 3) # used only for display
plt.ion()
fig = plt.figure()
if not os.path.exists('output'):
os.makedirs('output')
out_file = 'output/townCentreOut.top'
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(use_dlib= use_dlibTracker) #create instance of the SORT tracker
if use_dlibTracker:
print "Dlib Correlation tracker activated!"
else:
print "Kalman tracker activated!"
with open(out_file, 'w') as f_out:
frames = detector.get_total_frames()
#frames = 795
for frame in range(0, frames): #frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame+1)
total_frames +=1
if total_frames < 10:
jpg_name = "00000%d.jpg" %(frame+1)
elif total_frames < 100:
jpg_name = "0000%d.jpg" %(frame+1)
elif total_frames < 1000:
jpg_name = "000%d.jpg" %(frame+1)
fn = '/home/coco/dataset/MOT16/MOT15-01/img1/'+ jpg_name # video frames are extracted to 'test/Pictures%d.jpg' with ffmpeg
img = io.imread(fn)
if (display):
ax1 = fig.add_subplot(111, aspect='equal')
ax1.imshow(img)
if(use_dlibTracker):
plt.title('Dlib Correlation Tracker')
else:
plt.title('Kalman Tracker')
start_time = time.time()
#print("frame")
#print(frame+1)
#print(detections)
trackers = tracker.update(detections,img)
cycle_time = time.time() - start_time
total_time += cycle_time
print('frame: %d...took: %3fs'%(frame,cycle_time))
print("valid trackers")
print(len(trackers))
for d in trackers:
f_out.write('%d,%d,%d,%d,x,x,x,x,%.3f,%.3f,%.3f,%.3f\n' % (d[4], frame, 1, 1, d[0], d[1], d[2], d[3]))
if (display):
d = d.astype(np.int32)
ax1.add_patch(patches.Rectangle((d[0], d[1]), d[2] - d[0], d[3] - d[1], fill=False, lw=3,
ec=colours[d[4] % 32, :]))
print("frame")
print(frame+1)
print("draw tracker in the output image")
print(d[0])
print(d[1])
print(d[2])
print(d[3])
ax1.set_adjustable('box-forced')
#label
ax1.annotate('id = %d' % (d[4]), xy=(d[0], d[1]), xytext=(d[0], d[1]))
if detections != []:#detector is active in this frame
ax1.annotate(" DETECTOR", xy=(5, 45), xytext=(5, 45))
#plt.axis("off")
#fig.canvas.flush_events()
#plt.draw()
#fig.tight_layout()
#fig.savefig("/home/coco/dataset/results/frameOut/%s"%(save_name), dpi=200)
#axl.cla()
if (display):
plt.axis('off')
fig.canvas.flush_events()
plt.draw()
fig.tight_layout()
if(True):
fig.savefig("/home/coco/dataset/results/frameOut/%s"%(jpg_name),dpi = 200)
ax1.cla()
def main():
args = parse_args()
display = args.display
use_dlibTracker = args.use_dlibTracker
saver = args.saver
total_time = 0.0
total_frames = 0
# for disp
if display:
colours = np.random.rand(32, 3) # used only for display
plt.ion()
fig = plt.figure()
if not os.path.exists('output'):
os.makedirs('output')
out_file = 'output/townCentreOut.top'
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(
use_dlib=use_dlibTracker) #create instance of the SORT tracker
if use_dlibTracker:
print "Dlib Correlation tracker activated!"
else:
print "Kalman tracker activated!"
with open(out_file, 'w') as f_out:
frames = detector.get_total_frames()
for frame in range(0, frames): #frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame)
total_frames += 1
fn = 'test/frame%d.jpg' % (
frame + 1
) # video frames are extracted to 'test/Pictures%d.jpg' with ffmpeg
img = io.imread(fn)
if (display):
ax1 = fig.add_subplot(111, aspect='equal')
ax1.imshow(img)
if (use_dlibTracker):
plt.title('Dlib Correlation Tracker')
else:
plt.title('Kalman Tracker')
start_time = time.time()
#update tracker
trackers = tracker.update(detections, img)
cycle_time = time.time() - start_time
total_time += cycle_time
print('frame: %d...took: %3fs' % (frame, cycle_time))
for d in trackers:
f_out.write('%d,%d,%d,%d,x,x,x,x,%.3f,%.3f,%.3f,%.3f\n' %
(d[4], frame, 1, 1, d[0], d[1], d[2], d[3]))
if (display):
d = d.astype(np.int32)
ax1.add_patch(
patches.Rectangle((d[0], d[1]),
d[2] - d[0],
d[3] - d[1],
fill=False,
lw=3,
ec=colours[d[4] % 32, :]))
ax1.set_adjustable('box-forced')
#label
ax1.annotate('id = %d' % (d[4]),
xy=(d[0], d[1]),
xytext=(d[0], d[1]))
if detections != []: #detector is active in this frame
ax1.annotate(" DETECTOR", xy=(5, 45), xytext=(5, 45))
if (display):
plt.axis('off')
fig.canvas.flush_events()
plt.draw()
fig.tight_layout()
#save the frame with tracking boxes
if (saver):
fig.savefig("./frameOut/f%d.jpg" % (frame + 1), dpi=200)
ax1.cla()
print("Total Tracking took: %.3f for %d frames or %.1f FPS" %
(total_time, total_frames, total_frames / total_time))
def main():
args = parse_args()
display = args.display
use_dlibTracker = args.use_dlibTracker
saver = args.saver
total_time = 0.0
total_frames = 0
count = 0
cap = cv2.VideoCapture('ppp.avi')
# Check if camera opened successfully
if (cap.isOpened() == False):
print("Error opening video stream or file")
# Read until video is completed
while (cap.isOpened()):
# Capture frame-by-frame
ret, frame = cap.read()
if ret == True:
# Display the resulting frame
cv2.imshow('Frame', frame)
cv2.imwrite("test/Pictures%d.jpg" % count,
frame) # save frame as JPEG file
fn = 'test/Pictures%d.jpg' % (count + 1)
count += 1
# Press Q on keyboard to exit
if cv2.waitKey(25) & 0xFF == ord('q'):
break
# Break the loop
else:
break
# When everything done, release the video capture object
cap.release()
# Closes all the frames
cv2.destroyAllWindows()
# for disp
if display:
colours = np.random.rand(32, 3) # used only for display
plt.ion()
fig = plt.figure()
if not os.path.exists('output'):
os.makedirs('output')
out_file = 'output/townCentreOut.top'
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(
use_dlib=use_dlibTracker) #create instance of the SORT tracker
if use_dlibTracker:
print("Dlib Correlation tracker activated!")
else:
print("Kalman tracker activated!")
with open(out_file, 'w') as f_out:
print(detector.get_total_frames())
frames = detector.get_total_frames()
for frame in range(0, frames): #frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame)
total_frames += 1
fn = 'test/Pictures%d.jpg' % (
frame + 1
) # video frames are extracted to 'test/Pictures%d.jpg' with ffmpeg
img = io.imread(fn)
if (display):
ax1 = fig.add_subplot(111, aspect='equal')
ax1.imshow(img)
if (use_dlibTracker):
plt.title('Dlib Correlation Tracker')
else:
plt.title('Kalman Tracker')
start_time = time.time()
#update tracker
trackers = tracker.update(detections, img)
cycle_time = time.time() - start_time
total_time += cycle_time
print('frame: %d...took: %3fs' % (frame, cycle_time))
for d in trackers:
f_out.write('%d,%d,%d,%d,x,x,x,x,%.3f,%.3f,%.3f,%.3f\n' %
(d[4], frame, 1, 1, d[0], d[1], d[2], d[3]))
if (display):
d = d.astype(np.int32)
ax1.add_patch(
patches.Rectangle((d[0], d[1]),
d[2] - d[0],
d[3] - d[1],
fill=False,
lw=3,
ec=colours[d[4] % 32, :]))
ax1.set_adjustable('box-forced')
#label
ax1.annotate('id = %d' % (d[4]),
xy=(d[0], d[1]),
xytext=(d[0], d[1]))
if detections != []: #detector is active in this frame
ax1.annotate(" DETECTOR", xy=(5, 45), xytext=(5, 45))
if (display):
plt.axis('off')
fig.canvas.flush_events()
plt.draw()
fig.tight_layout()
#save the frame with tracking boxes
if (saver):
fig.savefig("./frameOut/f%d.jpg" % (frame + 1), dpi=200)
ax1.cla()
print("Total Tracking took: %.3f for %d frames or %.1f FPS" %
(total_time, total_frames, total_frames / total_time))
def main():
args = parse_args()
display = args.display
use_dlibTracker = args.use_dlibTracker
saver = args.saver
total_time = 0.0
total_frames = 0
# # for disp
# if display:
# colours = np.random.rand(32, 3) # used only for display
# plt.ion()
# fig = plt.figure()
if not os.path.exists('output'):
os.makedirs('output')
out_file = 'output/townCentreOut.top'
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(
use_dlib=use_dlibTracker) #create instance of the SORT tracker
if use_dlibTracker:
print("Dlib Correlation tracker activated!")
else:
print("Kalman tracker activated!")
fourcc = cv2.VideoWriter_fourcc(*'MP4V') #
# out = cv2.VideoWriter('test.mp4', fourcc, 30.0, (1920, 1080))
# cap = cv2.VideoCapture('/home/vdo-ubuntu/Videos/TownCentreXVID.avi')
out = cv2.VideoWriter('test.mp4', fourcc, 30.0, (960, 480))
cap = cv2.VideoCapture(
'/home/vdo-ubuntu/Videos/DVR_ch12_main_20180716130000_20180716140000_1.avi'
)
# out = cv2.VideoWriter('test.mp4', fourcc, 30.0, (640, 480))
# cap = cv2.VideoCapture(0)
with open(out_file, 'w') as f_out:
frames = detector.get_total_frames() #프레임 전체개수를 받음
# frames = 100
for frame in range(0, frames): #frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame)
total_frames += 1
success, img = cap.read()
# fn = 'data/Pictures%d.png' % (frame + 1) # video frames are extracted to 'test/Pictures%d.jpg' with ffmpeg
# img = io.imread(fn)
# if (display):
# ax1 = fig.add_subplot(111, aspect='equal')
# ax1.imshow(img)
# if(use_dlibTracker):
# plt.title('Dlib Correlation Tracker')
# else:
# plt.title('Kalman Tracker')
if success == True:
cv2.imshow('d', img)
start_time = time.time()
#update tracker
trackers = tracker.update(detections, img)
cycle_time = time.time() - start_time
total_time += cycle_time
print('frame: %d...took: %3fs' % (frame, cycle_time))
for d in trackers:
f_out.write('%d,%d,%d,%d,x,x,x,x,%.3f,%.3f,%.3f,%.3f\n' %
(d[4], frame, 1, 1, d[0], d[1], d[2], d[3]))
# if (display):
# d = d.astype(np.int32)
# ax1.add_patch(patches.Rectangle((d[0], d[1]), d[2] - d[0], d[3] - d[1], fill=False, lw=3,
# ec=colours[d[4] % 32, :]))
# ax1.set_adjustable('box-forced')
# #label
# ax1.annotate('id = %d' % (d[4]), xy=(d[0], d[1]), xytext=(d[0], d[1]))
# if detections != []:#detector is active in this frame
# ax1.annotate(" DETECTOR", xy=(5, 45), xytext=(5, 45))
cv2.rectangle(img, (int(d[0]), int(d[1])),
(int(d[2]), int(d[3])), (0, 255, 0), 2)
cv2.putText(img, 'id = %d' % (d[4]), (int(d[0]), int(d[3])),
cv2.FONT_HERSHEY_DUPLEX, 1, (0, 0, 255))
out.write(img)
# if (display):
# plt.axis('off')
# fig.canvas.flush_events()
# plt.draw()
# fig.tight_layout()
# #save the frame with tracking boxes
# if(saver):
# fig.savefig("./frameOut/f%d.jpg"%(frame+1),dpi = 200)
# ax1.cla()
print("Total Tracking took: %.3f for %d frames or %.1f FPS" %
(total_time, total_frames, total_frames / total_time))
total_time = 0.0
total_frames = 0
# for disp
if display:
colours = np.random.rand(32, 3) # used only for display
plt.ion()
fig = plt.figure()
if not os.path.exists('output'):
os.makedirs('output')
out_file = 'output/townCentreOut.top'
#init detector
detector = GroundTruthDetections()
#init tracker
tracker = Sort(use_dlib=use_dlibTracker) #create instance of the SORT tracker
if use_dlibTracker:
print("Dlib Correlation tracker activated!")
else:
print("Kalman tracker activated!")
with open(out_file, 'w') as f_out:
frames = detector.get_total_frames()
for frame in range(0, frames): #frame numbers begin at 0!
# get detections
detections = detector.get_detected_items(frame)