def cleanmosaic_video_byframe(opt, netG, netM):
path = opt.media_path
fps, imagepaths = video_init(opt, path)[:2]
positions = get_mosaic_positions(opt, netM, imagepaths, savemask=True)
# clean mosaic
for i, imagepath in enumerate(imagepaths, 0):
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = impro.imread(os.path.join('./tmp/video2image', imagepath))
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y - size:y + size, x - size:x + size]
if opt.traditional:
img_fake = runmodel.traditional_cleaner(img_mosaic, opt)
else:
img_fake = runmodel.run_pix2pix(img_mosaic, netG, opt)
mask = cv2.imread(os.path.join('./tmp/mosaic_mask', imagepath), 0)
img_result = impro.replace_mosaic(img_origin, img_fake, mask, x, y,
size, opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic', imagepath),
img_result)
print('\r', 'Clean Mosaic:' + str(i + 1) + '/' + str(len(imagepaths)))
ffmpeg.image2video(
fps, './tmp/replace_mosaic/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
def addmosaic_video(opt,netS):
path = opt.media_path
fps,imagepaths = video_init(opt,path)
# get position
positions = []
for i,imagepath in enumerate(imagepaths,1):
img = impro.imread(os.path.join('./tmp/video2image',imagepath))
mask,x,y,area = runmodel.get_ROI_position(img,netS,opt)
positions.append([x,y,area])
cv2.imwrite(os.path.join('./tmp/ROI_mask',imagepath),mask)
print('\r','Find ROI location:'+str(i)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print('\nOptimize ROI locations...')
mask_index = filt.position_medfilt(np.array(positions), 7)
# add mosaic
for i in range(len(imagepaths)):
mask = impro.imread(os.path.join('./tmp/ROI_mask',imagepaths[mask_index[i]]),'gray')
img = impro.imread(os.path.join('./tmp/video2image',imagepaths[i]))
if impro.mask_area(mask)>100:
img = mosaic.addmosaic(img, mask, opt)
cv2.imwrite(os.path.join('./tmp/addmosaic_image',imagepaths[i]),img)
print('\r','Add Mosaic:'+str(i+1)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print()
ffmpeg.image2video( fps,
'./tmp/addmosaic_image/output_%05d.'+opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,os.path.splitext(os.path.basename(path))[0]+'_add.mp4'))
def cleanmosaic_video_byframe(opt,netG,netM):
path = opt.media_path
fps,imagepaths = video_init(opt,path)
positions = []
# get position
for i,imagepath in enumerate(imagepaths,1):
img_origin = impro.imread(os.path.join('./tmp/video2image',imagepath))
x,y,size = runmodel.get_mosaic_position(img_origin,netM,opt)[:3]
positions.append([x,y,size])
print('\r','Find mosaic location:'+str(i)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print('\nOptimize mosaic locations...')
positions =np.array(positions)
for i in range(3):positions[:,i] = filt.medfilt(positions[:,i],opt.medfilt_num)
# clean mosaic
for i,imagepath in enumerate(imagepaths,0):
x,y,size = positions[i][0],positions[i][1],positions[i][2]
img_origin = impro.imread(os.path.join('./tmp/video2image',imagepath))
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y-size:y+size,x-size:x+size]
img_fake = runmodel.run_pix2pix(img_mosaic,netG,opt)
img_result = impro.replace_mosaic(img_origin,img_fake,x,y,size,opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic',imagepath),img_result)
print('\r','Clean Mosaic:'+str(i+1)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print()
ffmpeg.image2video( fps,
'./tmp/replace_mosaic/output_%05d.'+opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,os.path.splitext(os.path.basename(path))[0]+'_clean.mp4'))
def addmosaic_video(opt, netS):
path = opt.media_path
fps, imagepaths = video_init(opt, path)[:2]
length = len(imagepaths)
start_frame = int(imagepaths[0][7:13])
mask_index = get_roi_positions(opt, netS, imagepaths)[(start_frame - 1):]
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('preview', cv2.WINDOW_NORMAL)
# add mosaic
print('Step:3/4 -- Add Mosaic:')
t1 = time.time()
# print(mask_index)
for i, imagepath in enumerate(imagepaths, 1):
mask = impro.imread(
os.path.join(
opt.temp_dir + '/ROI_mask',
imagepaths[np.clip(mask_index[i - 1] - start_frame, 0,
1000000)]), 'gray')
img = impro.imread(
os.path.join(opt.temp_dir + '/video2image', imagepath))
if impro.mask_area(mask) > 100:
try: #Avoid unknown errors
img = mosaic.addmosaic(img, mask, opt)
except Exception as e:
print('Warning:', e)
t = Thread(target=cv2.imwrite,
args=(os.path.join(opt.temp_dir + '/addmosaic_image',
imagepath), img))
t.start()
os.remove(os.path.join(opt.temp_dir + '/video2image', imagepath))
#preview result and print
if not opt.no_preview:
cv2.imshow('preview', img)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i, length),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
print('Step:4/4 -- Convert images to video')
ffmpeg.image2video(
fps,
opt.temp_dir + '/addmosaic_image/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_add.mp4'))
def cleanmosaic_video_byframe(opt, netG, netM):
path = opt.media_path
fps, imagepaths = video_init(opt, path)[:2]
positions = get_mosaic_positions(opt, netM, imagepaths, savemask=True)
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('clean', cv2.WINDOW_NORMAL)
# clean mosaic
print('Step:3/4 -- Clean Mosaic:')
length = len(imagepaths)
for i, imagepath in enumerate(imagepaths, 0):
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = impro.imread(
os.path.join(opt.temp_dir + '/video2image', imagepath))
img_result = img_origin.copy()
if size > 100:
try: #Avoid unknown errors
img_mosaic = img_origin[y - size:y + size, x - size:x + size]
if opt.traditional:
img_fake = runmodel.traditional_cleaner(img_mosaic, opt)
else:
img_fake = runmodel.run_pix2pix(img_mosaic, netG, opt)
mask = cv2.imread(
os.path.join(opt.temp_dir + '/mosaic_mask', imagepath), 0)
img_result = impro.replace_mosaic(img_origin, img_fake, mask,
x, y, size, opt.no_feather)
except Exception as e:
print('Warning:', e)
cv2.imwrite(os.path.join(opt.temp_dir + '/replace_mosaic', imagepath),
img_result)
os.remove(os.path.join(opt.temp_dir + '/video2image', imagepath))
#preview result and print
if not opt.no_preview:
cv2.imshow('clean', img_result)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i + 1) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i + 1, len(imagepaths)),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
print('Step:4/4 -- Convert images to video')
ffmpeg.image2video(
fps,
opt.temp_dir + '/replace_mosaic/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
def cleanmosaic_video_fusion(opt,netG,netM):
path = opt.media_path
N = 25
INPUT_SIZE = 128
fps,imagepaths = video_init(opt,path)
positions = []
# get position
for i,imagepath in enumerate(imagepaths,1):
img_origin = impro.imread(os.path.join('./tmp/video2image',imagepath))
# x,y,size = runmodel.get_mosaic_position(img_origin,net_mosaic_pos,opt)[:3]
x,y,size,mask = runmodel.get_mosaic_position(img_origin,netM,opt)
cv2.imwrite(os.path.join('./tmp/mosaic_mask',imagepath), mask)
positions.append([x,y,size])
print('\r','Find mosaic location:'+str(i)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print('\nOptimize mosaic locations...')
positions =np.array(positions)
for i in range(3):positions[:,i] = filt.medfilt(positions[:,i],opt.medfilt_num)
# clean mosaic
for i,imagepath in enumerate(imagepaths,0):
x,y,size = positions[i][0],positions[i][1],positions[i][2]
img_origin = impro.imread(os.path.join('./tmp/video2image',imagepath))
mask = cv2.imread(os.path.join('./tmp/mosaic_mask',imagepath),0)
if size==0:
cv2.imwrite(os.path.join('./tmp/replace_mosaic',imagepath),img_origin)
else:
mosaic_input = np.zeros((INPUT_SIZE,INPUT_SIZE,3*N+1), dtype='uint8')
for j in range(0,N):
img = impro.imread(os.path.join('./tmp/video2image',imagepaths[np.clip(i+j-12,0,len(imagepaths)-1)]))
img = img[y-size:y+size,x-size:x+size]
img = impro.resize(img,INPUT_SIZE)
mosaic_input[:,:,j*3:(j+1)*3] = img
mask = impro.resize(mask,np.min(img_origin.shape[:2]))
mask = mask[y-size:y+size,x-size:x+size]
mask = impro.resize(mask, INPUT_SIZE)
mosaic_input[:,:,-1] = mask
mosaic_input = data.im2tensor(mosaic_input,bgr2rgb=False,use_gpu=opt.use_gpu,use_transform = False,is0_1 = False)
unmosaic_pred = netG(mosaic_input)
#unmosaic_pred = (unmosaic_pred.cpu().detach().numpy()*255)[0]
#img_fake = unmosaic_pred.transpose((1, 2, 0))
img_fake = data.tensor2im(unmosaic_pred,rgb2bgr = False ,is0_1 = False)
img_result = impro.replace_mosaic(img_origin,img_fake,x,y,size,opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic',imagepath),img_result)
print('\r','Clean Mosaic:'+str(i+1)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print()
ffmpeg.image2video( fps,
'./tmp/replace_mosaic/output_%05d.'+opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,os.path.splitext(os.path.basename(path))[0]+'_clean.mp4'))
def styletransfer_video(opt,netG):
path = opt.media_path
positions = []
fps,imagepaths = video_init(opt,path)
for i,imagepath in enumerate(imagepaths,1):
img = impro.imread(os.path.join('./tmp/video2image',imagepath))
img = runmodel.run_styletransfer(opt, netG, img)
cv2.imwrite(os.path.join('./tmp/style_transfer',imagepath),img)
print('\r','Transfer:'+str(i)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=40),end='')
print()
suffix = os.path.basename(opt.model_path).replace('.pth','').replace('style_','')
ffmpeg.image2video( fps,
'./tmp/style_transfer/output_%05d.'+opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,os.path.splitext(os.path.basename(path))[0]+'_'+suffix+'.mp4'))
def cleanmosaic_video_fusion(opt,netG,netM):
path = opt.media_path
N = 25
if 'HD' in os.path.basename(opt.model_path):
INPUT_SIZE = 256
else:
INPUT_SIZE = 128
fps,imagepaths,height,width = video_init(opt,path)
positions = get_mosaic_positions(opt,netM,imagepaths,savemask=True)
# clean mosaic
img_pool = np.zeros((height,width,3*N), dtype='uint8')
for i,imagepath in enumerate(imagepaths,0):
x,y,size = positions[i][0],positions[i][1],positions[i][2]
# image read stream
mask = cv2.imread(os.path.join('./tmp/mosaic_mask',imagepath),0)
if i==0 :
for j in range(0,N):
img_pool[:,:,j*3:(j+1)*3] = impro.imread(os.path.join('./tmp/video2image',imagepaths[np.clip(i+j-12,0,len(imagepaths)-1)]))
else:
img_pool[:,:,0:(N-1)*3] = img_pool[:,:,3:N*3]
img_pool[:,:,(N-1)*3:] = impro.imread(os.path.join('./tmp/video2image',imagepaths[np.clip(i+12,0,len(imagepaths)-1)]))
img_origin = img_pool[:,:,int((N-1)/2)*3:(int((N-1)/2)+1)*3]
if size==0: # can not find mosaic,
cv2.imwrite(os.path.join('./tmp/replace_mosaic',imagepath),img_origin)
else:
mosaic_input = np.zeros((INPUT_SIZE,INPUT_SIZE,3*N+1), dtype='uint8')
mosaic_input[:,:,0:N*3] = impro.resize(img_pool[y-size:y+size,x-size:x+size,:], INPUT_SIZE)
mask_input = impro.resize(mask,np.min(img_origin.shape[:2]))[y-size:y+size,x-size:x+size]
mosaic_input[:,:,-1] = impro.resize(mask_input, INPUT_SIZE)
mosaic_input = data.im2tensor(mosaic_input,bgr2rgb=False,use_gpu=opt.use_gpu,use_transform = False,is0_1 = False)
unmosaic_pred = netG(mosaic_input)
img_fake = data.tensor2im(unmosaic_pred,rgb2bgr = False ,is0_1 = False)
img_result = impro.replace_mosaic(img_origin,img_fake,mask,x,y,size,opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic',imagepath),img_result)
print('\r','Clean Mosaic:'+str(i+1)+'/'+str(len(imagepaths)),util.get_bar(100*i/len(imagepaths),num=35),end='')
print()
ffmpeg.image2video( fps,
'./tmp/replace_mosaic/output_%05d.'+opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,os.path.splitext(os.path.basename(path))[0]+'_clean.mp4'))
def cleanmosaic_video_byframe(opt):
netG = loadmodel.pix2pix(opt)
net_mosaic_pos = loadmodel.unet_clean(opt)
path = opt.media_path
util.clean_tempfiles()
fps = ffmpeg.get_video_infos(path)[0]
ffmpeg.video2voice(path, './tmp/voice_tmp.mp3')
ffmpeg.video2image(path,
'./tmp/video2image/output_%05d.' + opt.tempimage_type)
positions = []
imagepaths = os.listdir('./tmp/video2image')
imagepaths.sort()
# get position
for imagepath in imagepaths:
img_origin = impro.imread(os.path.join('./tmp/video2image', imagepath))
x, y, size = runmodel.get_mosaic_position(img_origin, net_mosaic_pos,
opt)[:3]
positions.append([x, y, size])
print('Find mosaic location:', imagepath)
print('Optimize mosaic locations...')
positions = np.array(positions)
for i in range(3):
positions[:, i] = filt.medfilt(positions[:, i], opt.medfilt_num)
# clean mosaic
for i, imagepath in enumerate(imagepaths, 0):
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = impro.imread(os.path.join('./tmp/video2image', imagepath))
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y - size:y + size, x - size:x + size]
img_fake = runmodel.run_pix2pix(img_mosaic, netG, opt)
img_result = impro.replace_mosaic(img_origin, img_fake, x, y, size,
opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic', imagepath),
img_result)
print('Clean Mosaic:', imagepath)
ffmpeg.image2video(
fps, './tmp/replace_mosaic/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
def styletransfer_video(opt, netG):
path = opt.media_path
positions = []
fps, imagepaths = video_init(opt, path)[:2]
print('Step:2/4 -- Transfer')
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('preview', cv2.WINDOW_NORMAL)
length = len(imagepaths)
for i, imagepath in enumerate(imagepaths, 1):
img = impro.imread(
os.path.join(opt.temp_dir + '/video2image', imagepath))
img = runmodel.run_styletransfer(opt, netG, img)
cv2.imwrite(os.path.join(opt.temp_dir + '/style_transfer', imagepath),
img)
os.remove(os.path.join(opt.temp_dir + '/video2image', imagepath))
#preview result and print
if not opt.no_preview:
cv2.imshow('preview', img)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i, len(imagepaths)),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
suffix = os.path.basename(opt.model_path).replace('.pth', '').replace(
'style_', '')
print('Step:4/4 -- Convert images to video')
ffmpeg.image2video(
fps,
opt.temp_dir + '/style_transfer/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_' + suffix +
'.mp4'))
def addmosaic_video(opt):
net = loadmodel.unet(opt)
path = opt.media_path
util.clean_tempfiles()
fps = ffmpeg.get_video_infos(path)[0]
ffmpeg.video2voice(path, './tmp/voice_tmp.mp3')
ffmpeg.video2image(path,
'./tmp/video2image/output_%05d.' + opt.tempimage_type)
imagepaths = os.listdir('./tmp/video2image')
imagepaths.sort()
# get position
positions = []
for imagepath in imagepaths:
print('Find ROI location:', imagepath)
img = impro.imread(os.path.join('./tmp/video2image', imagepath))
mask, x, y, area = runmodel.get_ROI_position(img, net, opt)
positions.append([x, y, area])
cv2.imwrite(os.path.join('./tmp/ROI_mask', imagepath), mask)
print('Optimize ROI locations...')
mask_index = filt.position_medfilt(np.array(positions), 7)
# add mosaic
print('Add mosaic to images...')
for i in range(len(imagepaths)):
mask = impro.imread(
os.path.join('./tmp/ROI_mask', imagepaths[mask_index[i]]))
img = impro.imread(os.path.join('./tmp/video2image', imagepaths[i]))
img = mosaic.addmosaic(img, mask, opt)
cv2.imwrite(os.path.join('./tmp/addmosaic_image', imagepaths[i]), img)
ffmpeg.image2video(
fps, './tmp/addmosaic_image/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_add.mp4'))
def cleanmosaic_video_fusion(opt, netG, netM):
path = opt.media_path
N, T, S = 2, 5, 3
LEFT_FRAME = (N * S)
POOL_NUM = LEFT_FRAME * 2 + 1
INPUT_SIZE = 256
FRAME_POS = np.linspace(0, (T - 1) * S, T, dtype=np.int64)
img_pool = []
previous_frame = None
init_flag = True
fps, imagepaths, height, width = video_init(opt, path)
positions = get_mosaic_positions(opt, netM, imagepaths, savemask=True)
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('clean', cv2.WINDOW_NORMAL)
# clean mosaic
print('Step:3/4 -- Clean Mosaic:')
length = len(imagepaths)
for i, imagepath in enumerate(imagepaths, 0):
x, y, size = positions[i][0], positions[i][1], positions[i][2]
input_stream = []
# image read stream
if i == 0: # init
for j in range(POOL_NUM):
img_pool.append(
impro.imread(
os.path.join(
opt.temp_dir + '/video2image',
imagepaths[np.clip(i + j - LEFT_FRAME, 0,
len(imagepaths) - 1)])))
else: # load next frame
img_pool.pop(0)
img_pool.append(
impro.imread(
os.path.join(
opt.temp_dir + '/video2image',
imagepaths[np.clip(i + LEFT_FRAME, 0,
len(imagepaths) - 1)])))
img_origin = img_pool[LEFT_FRAME]
img_result = img_origin.copy()
if size > 50:
try: #Avoid unknown errors
for pos in FRAME_POS:
input_stream.append(
impro.resize(
img_pool[pos][y - size:y + size,
x - size:x + size],
INPUT_SIZE)[:, :, ::-1])
if init_flag:
init_flag = False
previous_frame = input_stream[N]
previous_frame = data.im2tensor(previous_frame,
bgr2rgb=True,
gpu_id=opt.gpu_id)
input_stream = np.array(input_stream).reshape(
1, T, INPUT_SIZE, INPUT_SIZE, 3).transpose((0, 4, 1, 2, 3))
input_stream = data.to_tensor(data.normalize(input_stream),
gpu_id=opt.gpu_id)
with torch.no_grad():
unmosaic_pred = netG(input_stream, previous_frame)
img_fake = data.tensor2im(unmosaic_pred, rgb2bgr=True)
previous_frame = unmosaic_pred
# previous_frame = data.tensor2im(unmosaic_pred,rgb2bgr = True)
mask = cv2.imread(
os.path.join(opt.temp_dir + '/mosaic_mask', imagepath), 0)
img_result = impro.replace_mosaic(img_origin, img_fake, mask,
x, y, size, opt.no_feather)
except Exception as e:
init_flag = True
print('Error:', e)
else:
init_flag = True
cv2.imwrite(os.path.join(opt.temp_dir + '/replace_mosaic', imagepath),
img_result)
os.remove(os.path.join(opt.temp_dir + '/video2image', imagepath))
#preview result and print
if not opt.no_preview:
cv2.imshow('clean', img_result)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i + 1) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i + 1, len(imagepaths)),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
print('Step:4/4 -- Convert images to video')
ffmpeg.image2video(
fps,
opt.temp_dir + '/replace_mosaic/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
def cleanmosaic_video_fusion(opt, netG, netM):
path = opt.media_path
N = 25
if 'HD' in os.path.basename(opt.model_path):
INPUT_SIZE = 256
else:
INPUT_SIZE = 128
fps, imagepaths, height, width = video_init(opt, path)
positions = get_mosaic_positions(opt, netM, imagepaths, savemask=True)
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('clean', cv2.WINDOW_NORMAL)
# clean mosaic
print('Clean Mosaic:')
length = len(imagepaths)
img_pool = np.zeros((height, width, 3 * N), dtype='uint8')
mosaic_input = np.zeros((INPUT_SIZE, INPUT_SIZE, 3 * N + 1), dtype='uint8')
for i, imagepath in enumerate(imagepaths, 0):
x, y, size = positions[i][0], positions[i][1], positions[i][2]
# image read stream
mask = cv2.imread(
os.path.join(opt.temp_dir + '/mosaic_mask', imagepath), 0)
if i == 0:
for j in range(0, N):
img_pool[:, :, j * 3:(j + 1) * 3] = impro.imread(
os.path.join(
opt.temp_dir + '/video2image',
imagepaths[np.clip(i + j - 12, 0,
len(imagepaths) - 1)]))
else:
img_pool[:, :, 0:(N - 1) * 3] = img_pool[:, :, 3:N * 3]
img_pool[:, :, (N - 1) * 3:] = impro.imread(
os.path.join(
opt.temp_dir + '/video2image',
imagepaths[np.clip(i + 12, 0,
len(imagepaths) - 1)]))
img_origin = img_pool[:, :,
int((N - 1) / 2) * 3:(int((N - 1) / 2) + 1) * 3]
img_result = img_origin.copy()
if size > 100:
try: #Avoid unknown errors
#reshape to network input shape
mosaic_input[:, :, 0:N * 3] = impro.resize(
img_pool[y - size:y + size, x - size:x + size, :],
INPUT_SIZE)
mask_input = impro.resize(mask, np.min(
img_origin.shape[:2]))[y - size:y + size,
x - size:x + size]
mosaic_input[:, :, -1] = impro.resize(mask_input, INPUT_SIZE)
mosaic_input_tensor = data.im2tensor(mosaic_input,
bgr2rgb=False,
use_gpu=opt.use_gpu,
use_transform=False,
is0_1=False)
unmosaic_pred = netG(mosaic_input_tensor)
img_fake = data.tensor2im(unmosaic_pred,
rgb2bgr=False,
is0_1=False)
img_result = impro.replace_mosaic(img_origin, img_fake, mask,
x, y, size, opt.no_feather)
except Exception as e:
print('Warning:', e)
cv2.imwrite(os.path.join(opt.temp_dir + '/replace_mosaic', imagepath),
img_result)
#preview result and print
if not opt.no_preview:
cv2.imshow('clean', img_result)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i + 1) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i + 1, len(imagepaths)),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
ffmpeg.image2video(
fps,
opt.temp_dir + '/replace_mosaic/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
x, y, size = get_mosaic_position(img_origin)
positions.append([x, y, size])
print('Find Positions:', imagepath)
positions = np.array(positions)
for i in range(3):
positions[:, i] = signal.medfilt(positions[:, i], opt.medfilt_num)
for i, imagepath in enumerate(imagepaths, 0):
imagepath = os.path.join('./tmp/video2image', imagepath)
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = cv2.imread(imagepath)
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y - size:y + size, x - size:x + size]
img_fake = runmodel.run_pix2pix(img_mosaic,
netG,
use_gpu=opt.use_gpu)
img_result = replace_mosaic(img_origin, img_fake, x, y, size)
cv2.imwrite(
os.path.join('./tmp/replace_mosaic',
os.path.basename(imagepath)), img_result)
print('Clean Mosaic:', imagepath)
ffmpeg.image2video(
fps, './tmp/replace_mosaic/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] +
'_CleanMosaic.mp4'))
if musicinfos['freq'][i][j] > 260:
imgs = electric_imgs
else:
imgs = hit_imgs
for x in range(len(imgs)):
img = cv2.imread(
os.path.join('./dataset/mosquito/imgs/output',
'%05d' % (x + img_cnt) + '.jpg'))
if j == 0:
_img = cv2.resize(imgs[x], (1440, 810))
img[:810, :1440] = _img
if j == 1:
_img = cv2.resize(imgs[x], (480, 270))
img[0:270, 1440:1920] = _img
if j == 2:
_img = cv2.resize(imgs[x], (480, 270))
img[270:540, 1440:1920] = _img
if j == 3:
_img = cv2.resize(imgs[x], (480, 270))
img[540:810, 1440:1920] = _img
cv2.imwrite(
os.path.join('./dataset/mosquito/imgs/output',
'%05d' % (x + img_cnt) + '.jpg'), img)
except Exception as e:
print(e)
ffmpeg.image2video(60,
os.path.join('./dataset/mosquito/imgs/output', '%05d.jpg'),
'./tmp/test_output.wav', './tmp/result.mp4')
def cleanmosaic_video_fusion(opt):
net = loadmodel.pix2pix(opt)
net_mosaic_pos = loadmodel.unet_clean(opt)
path = opt.media_path
N = 25
util.clean_tempfiles()
fps = ffmpeg.get_video_infos(path)[0]
ffmpeg.video2voice(path, './tmp/voice_tmp.mp3')
ffmpeg.video2image(path,
'./tmp/video2image/output_%05d.' + opt.tempimage_type)
positions = []
imagepaths = os.listdir('./tmp/video2image')
imagepaths.sort()
# get position
for imagepath in imagepaths:
img_origin = impro.imread(os.path.join('./tmp/video2image', imagepath))
# x,y,size = runmodel.get_mosaic_position(img_origin,net_mosaic_pos,opt)[:3]
x, y, size, mask = runmodel.get_mosaic_position(
img_origin, net_mosaic_pos, opt)
cv2.imwrite(os.path.join('./tmp/mosaic_mask', imagepath), mask)
positions.append([x, y, size])
print('Find mosaic location:', imagepath)
print('Optimize mosaic locations...')
positions = np.array(positions)
for i in range(3):
positions[:, i] = filt.medfilt(positions[:, i], opt.medfilt_num)
# clean mosaic
print('Clean mosaic...')
for i, imagepath in enumerate(imagepaths, 0):
print('Clean mosaic:', imagepath)
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = impro.imread(os.path.join('./tmp/video2image', imagepath))
mask = cv2.imread(os.path.join('./tmp/mosaic_mask', imagepath), 0)
if size == 0:
cv2.imwrite(os.path.join('./tmp/replace_mosaic', imagepath),
img_origin)
else:
mosaic_input = np.zeros((256, 256, 3 * N + 1), dtype='uint8')
for j in range(0, N):
img = impro.imread(
os.path.join(
'./tmp/video2image',
imagepaths[np.clip(i + j - 12, 0,
len(imagepaths) - 1)]))
img = img[y - size:y + size, x - size:x + size]
img = impro.resize(img, 256)
mosaic_input[:, :, j * 3:(j + 1) * 3] = img
mask = impro.resize(mask, np.min(img_origin.shape[:2]))
mask = mask[y - size:y + size, x - size:x + size]
mask = impro.resize(mask, 256)
mosaic_input[:, :, -1] = mask
mosaic_input = data.im2tensor(mosaic_input,
bgr2rgb=False,
use_gpu=opt.use_gpu,
use_transform=False)
unmosaic_pred = net(mosaic_input)
unmosaic_pred = (unmosaic_pred.cpu().detach().numpy() * 255)[0]
img_fake = unmosaic_pred.transpose((1, 2, 0))
img_result = impro.replace_mosaic(img_origin, img_fake, x, y, size,
opt.no_feather)
cv2.imwrite(os.path.join('./tmp/replace_mosaic', imagepath),
img_result)
ffmpeg.image2video(
fps, './tmp/replace_mosaic/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_clean.mp4'))
def addmosaic_video(opt, netS):
path = opt.media_path
fps, imagepaths = video_init(opt, path)[:2]
length = len(imagepaths)
# get position
positions = []
t1 = time.time()
if not opt.no_preview:
cv2.namedWindow('preview', cv2.WINDOW_NORMAL)
print('Step:2/4 -- Find ROI location')
for i, imagepath in enumerate(imagepaths, 1):
img = impro.imread(
os.path.join(opt.temp_dir + '/video2image', imagepath))
mask, x, y, size, area = runmodel.get_ROI_position(img, netS, opt)
positions.append([x, y, area])
cv2.imwrite(os.path.join(opt.temp_dir + '/ROI_mask', imagepath), mask)
#preview result and print
if not opt.no_preview:
cv2.imshow('preview', mask)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i, length),
end='')
print('\nOptimize ROI locations...')
mask_index = filt.position_medfilt(np.array(positions), 7)
# add mosaic
print('Step:3/4 -- Add Mosaic:')
t1 = time.time()
for i, imagepath in enumerate(imagepaths, 1):
mask = impro.imread(
os.path.join(opt.temp_dir + '/ROI_mask',
imagepaths[mask_index[i - 1]]), 'gray')
img = impro.imread(
os.path.join(opt.temp_dir + '/video2image', imagepath))
if impro.mask_area(mask) > 100:
try: #Avoid unknown errors
img = mosaic.addmosaic(img, mask, opt)
except Exception as e:
print('Warning:', e)
cv2.imwrite(os.path.join(opt.temp_dir + '/addmosaic_image', imagepath),
img)
os.remove(os.path.join(opt.temp_dir + '/video2image', imagepath))
#preview result and print
if not opt.no_preview:
cv2.imshow('preview', img)
cv2.waitKey(1) & 0xFF
t2 = time.time()
print('\r',
str(i) + '/' + str(length),
util.get_bar(100 * i / length, num=35),
util.counttime(t1, t2, i, length),
end='')
print()
if not opt.no_preview:
cv2.destroyAllWindows()
print('Step:4/4 -- Convert images to video')
ffmpeg.image2video(
fps,
opt.temp_dir + '/addmosaic_image/output_%06d.' + opt.tempimage_type,
opt.temp_dir + '/voice_tmp.mp3',
os.path.join(opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_add.mp4'))
def main():
if opt.mode == 'add':
net = loadmodel.unet(opt)
path = opt.media_path
if util.is_img(path):
print('Add Mosaic:', path)
img = impro.imread(path)
mask = runmodel.get_ROI_position(img, net, opt)[0]
img = mosaic.addmosaic(img, mask, opt)
cv2.imwrite(os.path.join(opt.result_dir, os.path.basename(path)),
img)
elif util.is_video(path):
util.clean_tempfiles()
fps = ffmpeg.get_video_infos(path)[0]
ffmpeg.video2voice(path, './tmp/voice_tmp.mp3')
ffmpeg.video2image(
path, './tmp/video2image/output_%05d.' + opt.tempimage_type)
imagepaths = os.listdir('./tmp/video2image')
imagepaths.sort()
# get position
positions = []
for imagepath in imagepaths:
imagepath = os.path.join('./tmp/video2image', imagepath)
print('Find ROI location:', imagepath)
img = impro.imread(imagepath)
mask, x, y, area = runmodel.get_ROI_position(img, net, opt)
positions.append([x, y, area])
cv2.imwrite(
os.path.join('./tmp/ROI_mask',
os.path.basename(imagepath)), mask)
print('Optimize ROI locations...')
mask_index = filt.position_medfilt(np.array(positions), 7)
# add mosaic
print('Add mosaic to images...')
for i in range(len(imagepaths)):
mask_path = os.path.join('./tmp/ROI_mask',
imagepaths[mask_index[i]])
mask = impro.imread(mask_path)
img = impro.imread(
os.path.join('./tmp/video2image', imagepaths[i]))
img = mosaic.addmosaic(img, mask, opt)
cv2.imwrite(
os.path.join('./tmp/addmosaic_image',
os.path.basename(imagepaths[i])), img)
ffmpeg.image2video(
fps, './tmp/addmosaic_image/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] + '_add.mp4'))
elif opt.mode == 'clean':
netG = loadmodel.pix2pix(opt)
net_mosaic_pos = loadmodel.unet_clean(opt)
path = opt.media_path
if util.is_img(path):
print('Clean Mosaic:', path)
img_origin = impro.imread(path)
x, y, size = runmodel.get_mosaic_position(img_origin,
net_mosaic_pos, opt)
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y - size:y + size, x - size:x + size]
img_fake = runmodel.run_pix2pix(img_mosaic, netG, opt)
img_result = impro.replace_mosaic(img_origin, img_fake, x, y,
size, opt.no_feather)
cv2.imwrite(os.path.join(opt.result_dir, os.path.basename(path)),
img_result)
elif util.is_video(path):
util.clean_tempfiles()
fps = ffmpeg.get_video_infos(path)[0]
ffmpeg.video2voice(path, './tmp/voice_tmp.mp3')
ffmpeg.video2image(
path, './tmp/video2image/output_%05d.' + opt.tempimage_type)
positions = []
imagepaths = os.listdir('./tmp/video2image')
imagepaths.sort()
# get position
for imagepath in imagepaths:
imagepath = os.path.join('./tmp/video2image', imagepath)
img_origin = impro.imread(imagepath)
x, y, size = runmodel.get_mosaic_position(
img_origin, net_mosaic_pos, opt)
positions.append([x, y, size])
print('Find mosaic location:', imagepath)
print('Optimize mosaic locations...')
positions = np.array(positions)
for i in range(3):
positions[:, i] = filt.medfilt(positions[:, i],
opt.medfilt_num)
# clean mosaic
for i, imagepath in enumerate(imagepaths, 0):
imagepath = os.path.join('./tmp/video2image', imagepath)
x, y, size = positions[i][0], positions[i][1], positions[i][2]
img_origin = impro.imread(imagepath)
img_result = img_origin.copy()
if size != 0:
img_mosaic = img_origin[y - size:y + size,
x - size:x + size]
img_fake = runmodel.run_pix2pix(img_mosaic, netG, opt)
img_result = impro.replace_mosaic(img_origin, img_fake, x,
y, size, opt.no_feather)
cv2.imwrite(
os.path.join('./tmp/replace_mosaic',
os.path.basename(imagepath)), img_result)
print('Clean Mosaic:', imagepath)
ffmpeg.image2video(
fps, './tmp/replace_mosaic/output_%05d.' + opt.tempimage_type,
'./tmp/voice_tmp.mp3',
os.path.join(
opt.result_dir,
os.path.splitext(os.path.basename(path))[0] +
'_clean.mp4'))
util.clean_tempfiles(tmp_init=False)
fill_flags[start_frame + j] = match_index
img = cv2.imread(
os.path.join('./tmp/video_imgs', '%02d' % match_index,
'%05d' % (j + 1) + '.png'))
impro.imwrite(
os.path.join('./tmp/output_imgs',
'%05d' % (start_frame + j) + '.jpg'), img)
if i != (len(src_indexs) - 1) and start_frame + last_frame <= int(
np.round(src_indexs[i + 1] / 100 * FPS) - adv_frame):
for j in range(
int(np.round(src_indexs[i + 1] / 100 * FPS) - adv_frame) -
start_frame - last_frame + 1):
frame = start_frame + last_frame + j
fill_flags[frame] = match_index
img = cv2.imread(
os.path.join('./tmp/video_imgs', '%02d' % match_index,
'%05d' % (last_frame) + '.png'))
impro.imwrite(
os.path.join('./tmp/output_imgs',
'%05d' % (frame) + '.jpg'), img)
# print(fill_flags)
blackground = np.zeros((480, 640, 3), dtype=np.uint8)
for i in range(len(fill_flags)):
if fill_flags[i] == -1:
impro.imwrite(
os.path.join('./tmp/output_imgs', '%05d' % (i) + '.jpg'),
blackground)
ffmpeg.image2video(FPS, './tmp/output_imgs/%5d.jpg',
'./tmp/test_output.wav', './result.mp4')
index = np.argwhere(diff == np.min(diff))[0][0]
for x in range(len(imgs[index])):
frame_infos[j, int(fps * musicinfos['time'][i]) +
x] = np.array([index, x])
for i in range(frame_infos.shape[1]):
outimg = np.zeros((height, width, 3), dtype=np.uint8)
if frame_infos[0, i, 0] != -1:
outimg = imgs[frame_infos[0, i, 0]][frame_infos[0, i, 1]]
if showchord:
if frame_infos[1, i, 0] != -1:
outimg[height - 50 * 1 - height // 4 * 1:height - 50 * 1 -
height // 4 * 0, width - 50 * 1 - width // 4 * 1:width -
50 * 1] = resize_imgs[frame_infos[1, i,
0]][frame_infos[1, i, 1]]
if frame_infos[2, i, 0] != -1:
outimg[height - 50 * 2 - height // 4 * 2:height - 50 * 2 -
height // 4 * 1, width - 50 * 1 - width // 4 * 1:width -
50 * 1] = resize_imgs[frame_infos[1, i,
0]][frame_infos[1, i, 1]]
if frame_infos[3, i, 0] != -1:
outimg[height - 50 * 3 - height // 4 * 3:height - 50 * 3 -
height // 4 * 2, width - 50 * 1 - width // 4 * 1:width -
50 * 1] = resize_imgs[frame_infos[1, i,
0]][frame_infos[1, i, 1]]
cv2.imwrite(os.path.join('./tmp/output_img', '%05d' % i + '.jpg'), outimg)
ffmpeg.image2video(fps, os.path.join('./tmp/output_img', '%05d.jpg'),
'./tmp/test_output.wav', './tmp/result.mp4')
