def test_bgr2rgb(self):
in_img = np.random.rand(10, 10, 3).astype(np.float32)
out_img = mmcv.bgr2rgb(in_img)
assert out_img.shape == in_img.shape
assert_array_equal(out_img[..., 0], in_img[..., 2])
assert_array_equal(out_img[..., 1], in_img[..., 1])
assert_array_equal(out_img[..., 2], in_img[..., 0])
def show_result_pyplot(model,
img,
result,
score_thr=0.3,
fig_size=(15, 10),
out_file=None):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img,
result,
score_thr=score_thr,
show=False,
out_file=out_file)
if out_file is None:
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.show()
def show_result_pyplot(img,
result,
class_names,
score_thr=0.3,
fig_size=(15, 10),
out_file=None):
"""Visualize the detection results on the image.
Args:
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
class_names (list[str] or tuple[str]): A list of class names.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
out_file (str, optional): If specified, the visualization result will
be written to the out file instead of shown in a window.
"""
img = show_result(img,
result,
class_names,
score_thr=score_thr,
show=False,
out_file=out_file)
if out_file is None:
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
def main():
parser = ArgumentParser()
parser.add_argument('img_dir', help='Image dir')
parser.add_argument('config_dir', help='Config dir')
parser.add_argument('checkpoint_dir', help='Checkpoint dir')
parser.add_argument('result_dir', help='result dir')
parser.add_argument(
'--device', default='cuda:0', help='Device used for inference')
parser.add_argument(
'--score-thr', type=float, default=0.9, help='bbox score threshold')
args = parser.parse_args()
# build the model from a config file and a checkpoint file
model = init_detector(args.config_dir, args.checkpoint_dir, device=args.device)
# test some img
select_result = []
for img_file in os.listdir(args.img_dir):
print(img_file)
img = args.img_dir + img_file
result = inference_detector(model, img)
# show the results
result_pic = show_result_pyplot(model, img, result, score_thr=args.score_thr)
plt.figure(figsize=(15, 10))
plt.imshow(mmcv.bgr2rgb(result_pic))
plt.title('result')
plt.tight_layout()
plt.savefig(args.result_dir + img_file)
def save_result_pyplot(model,
img,
result,
output_path,
score_thr=0.3,
fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(
img,
result,
score_thr=score_thr,
show=False,
bbox_color='red',
text_color='red',
thickness=4,
font_scale=0.75,
)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.savefig(output_path, format='jpg')
plt.cla()
def show_result_pyplot(index,
model,
img,
result,
score_thr=0.3,
fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img,
result,
score_thr=score_thr,
show=False,
thickness=5)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.axis('off')
plt.savefig('./result_images/' + index[-16:],
bbox_inches='tight',
pad_inches=0)
plt.clf()
plt.close('all')
def show_result_pyplot(model,
img,
result,
score_thr=0.3,
fig_size=(15, 10),
title='result',
block=True):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
title (str): Title of the pyplot figure.
block (bool): Whether to block GUI.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, score_thr=score_thr, show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.title(title)
plt.tight_layout()
plt.show(block=block)
def show_pose_result(model,
img,
result,
kpt_score_thr=0.3,
skeleton=None,
fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str | np.ndarray): Image filename or loaded image.
result (list[dict]): The results to draw over `img`
(bbox_result, pose_result).
kpt_score_thr (float): The threshold to visualize the keypoints.
skeleton (list[tuple()]):
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img,
result,
skeleton,
kpt_score_thr=kpt_score_thr,
show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.show()
def show_result_pyplot(img: Union[str, np.ndarray],
result: np.ndarray,
palette: Optional[Iterable] = None,
fig_size: Iterable[int] = (15, 10),
opacity: float = 0.5,
title: str = '',
block: bool = True):
img = mmcv.imread(img)
img = img.copy()
seg = result[0]
seg = mmcv.imresize(seg, img.shape[:2][::-1])
palette = np.array(palette)
assert palette.shape[1] == 3
assert len(palette.shape) == 2
assert 0 < opacity <= 1.0
color_seg = np.zeros((seg.shape[0], seg.shape[1], 3), dtype=np.uint8)
for label, color in enumerate(palette):
color_seg[seg == label, :] = color
# convert to BGR
color_seg = color_seg[..., ::-1]
img = img * (1 - opacity) + color_seg * opacity
img = img.astype(np.uint8)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.title(title)
plt.tight_layout()
plt.show(block=block)
def _totensor(img, bgr2rgb, float32):
if img.shape[2] == 3 and bgr2rgb:
img = mmcv.bgr2rgb(img)
img = torch.from_numpy(img.transpose(2, 0, 1))
if float32:
img = img.float()
return img
def show_result_pyplot(model,
img_path,
result,
score_thr=0.3,
fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img, coordinates_list, label_txts = model.show_result(img_path,
result,
score_thr=score_thr,
show=False)
# color_scheme = extract_color_scheme(img, coordinates_list)
# color_scheme = scheme(img, coordinates_list)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.show()
# img = mmcv.bgr2rgb(img)
save_path = os.path.join('results', img_path.split('/')[-1])
mmcv.imwrite(img, save_path)
return img, coordinates_list, label_txts
def convert_color(img_group, to_grayscale, to_rgb):
if to_grayscale:
return [mmcv.bgr2gray(img, keepdim=True) for img in img_group]
elif to_rgb:
return [mmcv.bgr2rgb(img) for img in img_group]
else:
return img_group
def save_result_pyplot(img,
result,
class_names,
images_path,
results_path,
score_thr=0.3,
fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
class_names (list[str] or tuple[str]): A list of class names.
results_path: The path to the directory where the images should be saved
images_path: The path to the directory of the images
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
img_result = show_result(join(images_path, img),
result,
class_names,
score_thr=score_thr,
show=False)
plt.figure(figsize=fig_size)
plt.imsave(join(results_path, img), mmcv.bgr2rgb(img_result))
def get_display_img(args, item, pipelines):
"""get image to display."""
# srcs picture could be in RGB or BGR order due to different backends.
if args.bgr2rgb:
item['img'] = mmcv.bgr2rgb(item['img'])
src_image = item['img'].copy()
pipeline_images = [src_image]
# get intermediate images through pipelines
if args.mode in ['transformed', 'concat', 'pipeline']:
for pipeline in pipelines.values():
item = pipeline(item)
trans_image = copy.deepcopy(item['img'])
trans_image = np.ascontiguousarray(trans_image, dtype=np.uint8)
pipeline_images.append(trans_image)
# concatenate images to be showed according to mode
if args.mode == 'original':
image = prepare_imgs(args, [src_image], ['src'])
elif args.mode == 'transformed':
image = prepare_imgs(args, [pipeline_images[-1]], ['transformed'])
elif args.mode == 'concat':
steps = ['src', 'transformed']
image = prepare_imgs(args, [pipeline_images[0], pipeline_images[-1]],
steps)
elif args.mode == 'pipeline':
steps = ['src'] + list(pipelines.keys())
image = prepare_imgs(args, pipeline_images, steps)
return image
def show_result_pyplot(model,
img,
result,
palette=None,
fig_size=(15, 10),
opacity=0.5):
"""Visualize the segmentation results on the image.
Args:
model (nn.Module): The loaded segmentor.
img (str or np.ndarray): Image filename or loaded image.
result (list): The segmentation result.
palette (list[list[int]]] | None): The palette of segmentation
map. If None is given, random palette will be generated.
Default: None
fig_size (tuple): Figure size of the pyplot figure.
opacity(float): Opacity of painted segmentation map.
Default 0.5.
Must be in (0, 1] range.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img,
result,
palette=palette,
show=False,
opacity=opacity)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.show()
def get_image(self, image, transform_I):
# transform
img = mmcv.bgr2rgb(image)
crop_size = Image.fromarray(img).size
img = transform_I(Image.fromarray(img))
return img, crop_size
def imshow_bboxes_w_ids(img,
bboxes,
ids,
thickness=1,
font_scale=0.5,
out_file=None):
assert bboxes.ndim == 2
assert ids.ndim == 1
assert bboxes.shape[0] == ids.shape[0]
assert bboxes.shape[1] == 4 or bboxes.shape[1] == 5
if isinstance(img, str):
img = plt.imread(img)
else:
img = mmcv.bgr2rgb(img)
plt.imshow(img)
plt.gca().set_axis_off()
plt.autoscale(False)
plt.subplots_adjust(top=1,
bottom=0,
right=1,
left=0,
hspace=None,
wspace=None)
plt.margins(0, 0)
plt.gca().xaxis.set_major_locator(plt.NullLocator())
plt.gca().yaxis.set_major_locator(plt.NullLocator())
for bbox, id in zip(bboxes, ids):
bbox_int = bbox.astype(np.int32)
left_top = (bbox_int[0], bbox_int[1])
w = bbox_int[2] - bbox_int[0] + 1
h = bbox_int[3] - bbox_int[1] + 1
color = random_color(id)
plt.gca().add_patch(
Rectangle(left_top,
w,
h,
thickness,
edgecolor=color,
facecolor='none'))
label_text = '{}'.format(int(id))
bg_height = 15
bg_width = 12
bg_width = len(label_text) * bg_width
plt.gca().add_patch(
Rectangle((left_top[0], left_top[1] - bg_height),
bg_width,
bg_height,
thickness,
edgecolor=color,
facecolor=color))
plt.text(left_top[0] - 1, left_top[1], label_text, fontsize=5)
if out_file is not None:
plt.savefig(out_file, dpi=300, bbox_inches='tight', pad_inches=0.0)
plt.show()
plt.clf()
return img
def _totensor(img, bgr2rgb, float32):
if img.shape[2] == 3 and bgr2rgb:
img = mmcv.bgr2rgb(img)
#img = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)
#img = torch.from_numpy(img.transpose(2, 0, 1))
img = trans(img)
if float32:
img = img.float()
return img
def get_image(self, image, transform_I, size, crop_coords):
# crop
img = mmcv.bgr2rgb(image)
if self.opt.isTrain and not self.use_for_finetune:
img = self.crop(Image.fromarray(img), crop_coords)
else:
img = Image.fromarray(img)
crop_size = img.size
# transform
img = transform_I(img)
return img, crop_size
def get_image(self, image, transform_I, crop_coords):
# transform
img = mmcv.bgr2rgb(image)
img = self.crop(Image.fromarray(img), crop_coords) # debug
crop_size = img.size
# crop_size = Image.fromarray(img).size
# debug
# img = transform_I(Image.fromarray(img))
img = transform_I(img)
return img, crop_size
def show_ann(coco, img, ann_info):
"""Show the img.
:param coco: the class of coco
:type coco: object
:param img: image
:type img: ndarray
:param ann_info: the annotation information
:type ann_info: tuple
"""
plt.imshow(mmcv.bgr2rgb(img))
plt.axis('off')
coco.showAnns(ann_info)
plt.show()
def show_result_pyplot(model, img, result, fig_size=(15, 10)):
"""Visualize the classification results on the image.
Args:
model (nn.Module): The loaded classifier.
img (str or np.ndarray): Image filename or loaded image.
result (list): The classification result.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.show()
def show_result_pyplot(model, img_file, result, score_thr=0.3, fig_size=(16, 9)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img_file, result, score_thr=score_thr, show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
plt.savefig(img_file.split('.')[0] + '_det_results.jpg', dpi=120)
def get_result_image(img):
model = load_model()
result = inference_detector(model, img)
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, score_thr=0.3, show=False)
plt.figure(figsize=(10, 10))
plt.axis('off')
plt.imshow(mmcv.bgr2rgb(img))
plt.tight_layout()
now = datetime.now()
dt = now.strftime("%m%d%H%M%S")
result_img_path = './static/outputs/output_' + dt + '.png'
plt.savefig(result_img_path)
result_img_path = result_img_path[9:]
return result_img_path
def nondist_validation(self, dataloader, current_iter, tb_logger,
save_img):
assert dataloader is None, 'Validation dataloader should be None.'
self.test()
result = tensor2img(self.output, min_max=(-1, 1))
if self.opt['is_train']:
save_img_path = osp.join(self.opt['path']['visualization'],
'train', f'train_{current_iter}.png')
else:
save_img_path = osp.join(self.opt['path']['visualization'], 'test',
f'test_{self.opt["name"]}.png')
mmcv.imwrite(result, save_img_path)
# add sample images to tb_logger
result = (result / 255.).astype(np.float32)
result = mmcv.bgr2rgb(result)
if tb_logger is not None:
tb_logger.add_image(
'samples', result, global_step=current_iter, dataformats='HWC')
def show_result_pyplot(model, img, result, score_thr=0.3, fig_size=(15, 10)):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, score_thr=score_thr, show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
# plt.savefig(save_pth + "/{}.png".format(im_str.split('.')[0]), bbox_inches='tight')
plt.show()
def get_display_img(item, pipeline, mode, bgr2rgb):
"""get image to display."""
if bgr2rgb:
item['img'] = mmcv.bgr2rgb(item['img'])
src_image = item['img'].copy()
# get transformed picture
if mode in ['pipeline', 'concat']:
item = pipeline(item)
trans_image = item['img']
trans_image = np.ascontiguousarray(trans_image, dtype=np.uint8)
if mode == 'concat':
image = concat(src_image, trans_image)
elif mode == 'original':
image = src_image
elif mode == 'pipeline':
image = trans_image
return image
def vis3d():
v_path = os.path.join(root, 'unzip', 'test_video/id03862/jc6k4sbenMY/00366' + '.mp4' )
lmark_path = os.path.join(root, 'unzip', 'test_video/id03862/jc6k4sbenMY/00366_prnet' + '.npy' )
# lmark_path = os.path.join(root, 'unzip', 'test_video/id04276/k0zLls_oen0/00341_prnet' + '.npy' )
cap = cv2.VideoCapture(v_path)
lmark = np.load(lmark_path)
print (lmark.shape)
print (lmark[0])
count = 0
tmp = v_path.split('/')
real_video = mmcv.VideoReader(v_path)
for count in range(0,100):
input = real_video[count]
input = mmcv.bgr2rgb(input)
preds = lmark#[count]
#TODO: Make this nice
fig = plt.figure(figsize=plt.figaspect(.5))
ax = fig.add_subplot(1, 2, 1)
ax.imshow(input)
ax.plot(preds[0:17,0],preds[0:17,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[17:22,0],preds[17:22,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[22:27,0],preds[22:27,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[27:31,0],preds[27:31,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[31:36,0],preds[31:36,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[36:42,0],preds[36:42,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[42:48,0],preds[42:48,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[48:60,0],preds[48:60,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.plot(preds[60:68,0],preds[60:68,1],marker='o',markersize=1,linestyle='-',color='w',lw=1)
ax.axis('off')
ax = fig.add_subplot(1, 2, 2, projection='3d')
surf = ax.scatter(preds[:,0]*1.2,preds[:,1],preds[:,2],c="cyan", alpha=1.0, edgecolor='b')
ax.plot3D(preds[:17,0]*1.2,preds[:17,1], preds[:17,2], color='blue' )
ax.plot3D(preds[17:22,0]*1.2,preds[17:22,1],preds[17:22,2], color='blue')
ax.plot3D(preds[22:27,0]*1.2,preds[22:27,1],preds[22:27,2], color='blue')
ax.plot3D(preds[27:31,0]*1.2,preds[27:31,1],preds[27:31,2], color='blue')
ax.plot3D(preds[31:36,0]*1.2,preds[31:36,1],preds[31:36,2], color='blue')
ax.plot3D(preds[36:42,0]*1.2,preds[36:42,1],preds[36:42,2], color='blue')
ax.plot3D(preds[42:48,0]*1.2,preds[42:48,1],preds[42:48,2], color='blue')
ax.plot3D(preds[48:,0]*1.2,preds[48:,1],preds[48:,2], color='blue' )
ax.view_init(elev=90., azim=90.)
ax.set_xlim(ax.get_xlim()[::-1])
plt.show()
def show_result_pyplot(model, img, result, palette=None, fig_size=(15, 10)):
"""Visualize the segmentation results on the image.
Args:
model (nn.Module): The loaded segmentor.
img (str or np.ndarray): Image filename or loaded image.
result (list): The segmentation result.
palette (list[list[int]]] | None): The palette of segmentation
map. If None is given, random palette will be generated.
Default: None
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, palette=palette, show=False)
plt.figure(figsize=fig_size)
plt.imshow(mmcv.bgr2rgb(img))
#plt.imsave('my_demo.png', mmcv.bgr2rgb(img))
plt.savefig('my_demo.png')
def show_result_pyplot(model,
img,
result,
score_thr=0.3,
fig_size=(15, 10),
name=1):
"""Visualize the detection results on the image.
Args:
model (nn.Module): The loaded detector.
img (str or np.ndarray): Image filename or loaded image.
result (tuple[list] or list): The detection result, can be either
(bbox, segm) or just bbox.
score_thr (float): The threshold to visualize the bboxes and masks.
fig_size (tuple): Figure size of the pyplot figure.
"""
if hasattr(model, 'module'):
model = model.module
img = model.show_result(img, result, score_thr=score_thr, show=False)
# plt.figure(figsize=fig_size)
# plt.imshow(mmcv.bgr2rgb(img))
# import pdb; pdb.set_trace()
plt.imsave('/home/pengyi/temp{}.jpg'.format(name), mmcv.bgr2rgb(img))
plt.show()