def demo(opt):
result_root = opt.output_root if opt.output_root != '' else '.'
mkdir_if_missing(result_root)
logger.info('Starting tracking...')
#dataloader = datasets.LoadVideo(opt.input_video, opt.img_size)
dataloader = datasets.LoadImages(opt.input_video)
result_filename = os.path.join(result_root, 'results.txt')
frame_rate = 0.5 #dataloader.frame_rate
frame_dir = None if opt.output_format == 'text' else osp.join(
result_root, 'frame')
try:
eval_seq(opt,
dataloader,
'mot',
result_filename,
save_dir=frame_dir,
show_image=False,
frame_rate=frame_rate)
except Exception as e:
logger.info(e)
if opt.output_format == 'video':
output_video_path = osp.join(result_root, 'result.mp4')
cmd_str = 'ffmpeg -f image2 -i {}/%05d.jpg -b 5000k -c:v mpeg4 {}'.format(
osp.join(result_root, 'frame'), output_video_path)
os.system(cmd_str)
def run_demo(opt):
"""
:param opt:
:return:
"""
result_root = opt.output_root if opt.output_root != '' else '.'
mkdir_if_missing(result_root)
# clear existing frame results
frame_res_dir = result_root + '/frames'
if os.path.isdir(frame_res_dir):
shutil.rmtree(frame_res_dir)
os.makedirs(frame_res_dir)
else:
os.makedirs(frame_res_dir)
if opt.input_mode == 'video':
logger.info('Starting tracking...')
data_loader = datasets.LoadVideo(opt.input_video, opt.img_size)
elif opt.input_mode == 'image_dir':
logger.info('Starting detection...')
data_loader = datasets.LoadImages(opt.input_img, opt.img_size)
elif opt.input_mode == 'img_path_list_txt':
if not os.path.isfile(opt.input_img):
print('[Err]: invalid image file path list.')
return
with open(opt.input_img, 'r', encoding='utf-8') as r_h:
logger.info('Starting detection...')
paths = [x.strip() for x in r_h.readlines()]
print('Total {:d} image files.'.format(len(paths)))
data_loader = datasets.LoadImages(path=paths,
img_size=opt.img_size)
result_file_name = os.path.join(result_root, 'results.txt')
frame_rate = data_loader.frame_rate
frame_dir = None if opt.output_format == 'text' else osp.join(
result_root, 'frame')
opt.device = device
try: # 视频推断的入口函数
if opt.id_weight > 0:
eval_seq(opt=opt,
data_loader=data_loader,
data_type='mot',
result_f_name=result_file_name,
save_dir=frame_dir,
show_image=False,
frame_rate=frame_rate,
mode='track')
else:
# eval_seq(opt=opt,
# data_loader=data_loader,
# data_type='mot',
# result_f_name=result_file_name,
# save_dir=frame_dir,
# show_image=False,
# frame_rate=frame_rate,
# mode='detect')
# only for tmp detection evaluation...
output_dir = '/users/duanyou/c5/results_new/results_all/mcmot_hrnet18_deconv_ep3'
eval_seq_and_output_dets(opt=opt,
data_loader=data_loader,
data_type='mot',
result_f_name=result_file_name,
out_dir=output_dir,
save_dir=frame_dir,
show_image=False)
except Exception as e:
logger.info(e)
if opt.output_format == 'video':
output_video_path = osp.join(result_root, 'result.mp4')
cmd_str = 'ffmpeg -f image2 -i {}/%05d.jpg -b 5000k -c:v mpeg4 {}' \
.format(osp.join(result_root, 'frame'),
output_video_path)
os.system(cmd_str)
def run_demo(opt):
"""
:param opt:
:return:
"""
result_root = opt.output_root if opt.output_root != '' else '.'
mkdir_if_missing(result_root)
# clear existing frame results
frame_res_dir = result_root + '/frame'
if os.path.isdir(frame_res_dir):
shutil.rmtree(frame_res_dir)
os.makedirs(frame_res_dir)
else:
os.makedirs(frame_res_dir)
if opt.input_mode == 'video':
if opt.id_weight > 0:
logger.info('Starting tracking...')
else:
logger.info('Starting detection...')
if not os.path.isfile(opt.input_video):
print('[Err]: invalid input video file.')
return
data_loader = datasets.LoadVideo(opt.input_video,
opt.img_size) # load video as input
f_name = os.path.split(opt.input_video)[-1][:-4]
elif opt.input_mode == 'image_dir':
logger.info('Starting detection...')
data_loader = datasets.LoadImages(opt.input_img,
opt.img_size) # load images as input
f_name = os.path.split(opt.input_video)[-1]
opt.id_weight = 0 # only do detection in this mode
elif opt.input_mode == 'img_path_list_txt':
logger.info('Starting detection...')
if not os.path.isfile(opt.input_img):
print('[Err]: invalid image file path list.')
return
opt.id_weight = 0 # only do detection in this mode
with open(opt.input_img, 'r', encoding='utf-8') as r_h:
logger.info('Starting detection...')
paths = [x.strip() for x in r_h.readlines()]
print('Total {:d} image files.'.format(len(paths)))
data_loader = datasets.LoadImages(path=paths,
img_size=opt.img_size)
result_file_name = os.path.join(result_root, 'results.txt')
frame_rate = data_loader.frame_rate
frame_dir = None if opt.output_format == 'text' else osp.join(
result_root, 'frame')
# Set device
# opt.device = device
# set device
opt.device = str(FindFreeGPU())
print('Using gpu: {:s}'.format(opt.device))
device = select_device(
device='cpu' if not torch.cuda.is_available() else opt.device)
opt.device = device
try:
if opt.input_mode == 'video':
if opt.id_weight > 0:
eval_seq(opt=opt,
data_loader=data_loader,
data_type='mot',
result_f_name=result_file_name,
save_dir=frame_dir,
show_image=False,
frame_rate=frame_rate,
mode='track')
else: # input video, do detection
eval_seq(opt=opt,
data_loader=data_loader,
data_type='mot',
result_f_name=result_file_name,
save_dir=frame_dir,
show_image=False,
frame_rate=frame_rate,
mode='detect')
else:
# only for tmp detection evaluation...
output_dir = '/users/duanyou/c5/results_new/results_all/tmp'
eval_imgs_output_dets(opt=opt,
data_loader=data_loader,
data_type='mot',
result_f_name=result_file_name,
out_dir=output_dir,
save_dir=frame_dir,
show_image=False)
except Exception as e:
logger.info(e)
if opt.output_format == 'video':
output_video_path = 'result.mp4'
if opt.input_mode == 'video':
if opt.id_weight > 0:
output_video_path = osp.join(result_root,
f_name + '_track.mp4')
else:
output_video_path = osp.join(result_root, f_name + '_det.mp4')
elif opt.input_mode == 'image_dir':
output_video_path = osp.join(result_root, f_name + '_det.mp4')
cmd_str = 'ffmpeg -f image2 -i {}/%05d.jpg -b 5000k -c:v mpeg4 {}' \
.format(osp.join(result_root, 'frame'), output_video_path)
print(cmd_str)
os.system(cmd_str)
def main(opt,
data_root='/data/MOT16/train',
det_root=None,
seqs=('MOT16-05', ),
exp_name='demo',
save_images=False,
save_videos=False,
show_image=True):
"""
"""
logger.setLevel(logging.INFO)
result_root = os.path.join(data_root, '..', 'results', exp_name)
mkdir_if_missing(result_root)
data_type = 'mot'
# run tracking
accs = []
n_frame = 0
timer_avgs, timer_calls = [], []
for seq in seqs:
output_dir = os.path.join(data_root, '..', 'outputs', exp_name,
seq) if save_images or save_videos else None
logger.info('start seq: {}'.format(seq))
dataloader = datasets.LoadImages(osp.join(data_root, seq, 'img1'),
opt.img_size)
result_filename = os.path.join(result_root, '{}.txt'.format(seq))
meta_info = open(os.path.join(data_root, seq, 'seqinfo.ini')).read()
frame_rate = int(meta_info[meta_info.find('frameRate') +
10:meta_info.find('\nseqLength')])
nf, ta, tc = eval_seq(opt,
dataloader,
data_type,
result_filename,
save_dir=output_dir,
show_image=show_image,
frame_rate=frame_rate)
n_frame += nf
timer_avgs.append(ta)
timer_calls.append(tc)
# eval
logger.info('Evaluate seq: {}'.format(seq))
evaluator = Evaluator(data_root, seq, data_type)
accs.append(evaluator.eval_file(result_filename))
if save_videos:
output_video_path = osp.join(output_dir, '{}.mp4'.format(seq))
cmd_str = 'ffmpeg -f image2 -i {}/%05d.jpg -c:v copy {}'.format(
output_dir, output_video_path)
os.system(cmd_str)
timer_avgs = np.asarray(timer_avgs)
timer_calls = np.asarray(timer_calls)
all_time = np.dot(timer_avgs, timer_calls)
avg_time = all_time / np.sum(timer_calls)
logger.info('Time elapsed: {:.2f} seconds, FPS: {:.2f}'.format(
all_time, 1.0 / avg_time))
# get summary
metrics = mm.metrics.motchallenge_metrics
mh = mm.metrics.create()
summary = Evaluator.get_summary(accs, seqs, metrics)
strsummary = mm.io.render_summary(summary,
formatters=mh.formatters,
namemap=mm.io.motchallenge_metric_names)
print(strsummary)
Evaluator.save_summary(
summary, os.path.join(result_root, 'summary_{}.xlsx'.format(exp_name)))
