def __call__(self, img, boxes, labels, masks=None):
if self.color_choose == 0:
if random.uniform() < self.gray_p:
gray = mmcv.bgr2gray(img)
img = mmcv.gray2bgr(gray)
return img, boxes, labels, masks
# random brightness
if random.randint(2):
delta = random.uniform(-self.brightness_delta,
self.brightness_delta)
img += delta
# mode == 0 --> do random contrast first
# mode == 1 --> do random contrast last
mode = random.randint(2)
if mode == 1:
if random.randint(2):
alpha = random.uniform(self.contrast_lower,
self.contrast_upper)
img *= alpha
# convert color from BGR to HSV
img = mmcv.bgr2hsv(img)
# random saturation
if random.randint(2):
img[..., 1] *= random.uniform(self.saturation_lower,
self.saturation_upper)
# random hue
if random.randint(2):
img[..., 0] += random.uniform(-self.hue_delta, self.hue_delta)
img[..., 0][img[..., 0] > 360] -= 360
img[..., 0][img[..., 0] < 0] += 360
# convert color from HSV to BGR
img = mmcv.hsv2bgr(img)
# random contrast
if mode == 0:
if random.randint(2):
alpha = random.uniform(self.contrast_lower,
self.contrast_upper)
img *= alpha
# randomly swap channels
if random.randint(2):
img = img[..., random.permutation(3)]
else:
if self.color_choose == 1:
# random brightness
if random.randint(2):
delta = random.uniform(-self.brightness_delta,
self.brightness_delta)
img += delta
elif self.color_choose == 2:
# random contrast first
if random.randint(2):
alpha = random.uniform(self.contrast_lower,
self.contrast_upper)
img *= alpha
else:
# convert color from BGR to HSV
img = mmcv.bgr2hsv(img)
if self.color_choose == 3:
# random saturation
if random.randint(2):
img[..., 1] *= random.uniform(self.saturation_lower,
self.saturation_upper)
if self.color_choose == 4:
# random hue
if random.randint(2):
img[..., 0] += random.uniform(-self.hue_delta,
self.hue_delta)
img[..., 0][img[..., 0] > 360] -= 360
img[..., 0][img[..., 0] < 0] += 360
# convert color from HSV to BGR
img = mmcv.hsv2bgr(img)
return img, boxes, labels, masks
def process(self, inputs, outputs, out_dict):
"""
Args:
inputs: the inputs to a model.
It is a list of dict. Each dict corresponds to an image and
contains keys like "height", "width", "file_name", "image_id", "scene_id".
outputs: stores time
"""
cfg = self.cfg
if cfg.TEST.USE_PNP:
if cfg.TEST.PNP_TYPE.lower() == "ransac_pnp":
return self.process_pnp_ransac(inputs, outputs, out_dict)
elif cfg.TEST.PNP_TYPE.lower() == "net_iter_pnp":
return self.process_net_and_pnp(inputs,
outputs,
out_dict,
pnp_type="iter")
elif cfg.TEST.PNP_TYPE.lower() == "net_ransac_pnp":
return self.process_net_and_pnp(inputs,
outputs,
out_dict,
pnp_type="ransac")
else:
raise NotImplementedError
out_rots = out_dict["rot"].detach().to(self._cpu_device).numpy()
out_transes = out_dict["trans"].detach().to(self._cpu_device).numpy()
out_i = -1
for i, (_input, output) in enumerate(zip(inputs, outputs)):
json_results = []
start_process_time = time.perf_counter()
for inst_i in range(len(_input["roi_img"])):
out_i += 1 # the index in the flattened output
scene_im_id_split = _input["scene_im_id"][inst_i].split("/")
K = _input["cam"][inst_i].cpu().numpy().copy()
roi_label = _input["roi_cls"][inst_i] # 0-based label
score = _input["score"][inst_i]
roi_label, cls_name = self._maybe_adapt_label_cls_name(
roi_label)
if cls_name is None:
continue
# scene_id = int(scene_im_id_split[0])
scene_id = scene_im_id_split[0]
im_id = int(scene_im_id_split[1])
obj_id = self.data_ref.obj2id[cls_name]
# get pose
rot_est = out_rots[out_i]
trans_est = out_transes[out_i]
pose_est = np.hstack([rot_est, trans_est.reshape(3, 1)])
if cfg.DEBUG: # visualize pose
file_name = _input["file_name"][inst_i]
if f"{int(scene_id)}/{im_id}" != "9/47":
continue
im_ori = mmcv.imread(file_name, "color")
bbox = _input["bbox_est"][inst_i].cpu().numpy().copy()
im_vis = vis_image_bboxes_cv2(im_ori, [bbox],
[f"{cls_name}_{score}"])
self.ren.clear()
self.ren.draw_background(
mmcv.bgr2gray(im_ori, keepdim=True))
self.ren.draw_model(
self.ren_models[self.data_ref.objects.index(cls_name)],
pose_est)
ren_im, _ = self.ren.finish()
grid_show(
[ren_im[:, :, ::-1], im_vis[:, :, ::-1]],
[f"ren_im_{cls_name}", f"{scene_id}/{im_id}_{score}"],
row=1,
col=2,
)
json_results.extend(
self.pose_prediction_to_json(pose_est,
scene_id,
im_id,
obj_id=obj_id,
score=score,
pose_time=output["time"],
K=K))
output["time"] += time.perf_counter() - start_process_time
# process time for this image
for item in json_results:
item["time"] = output["time"]
self._predictions.extend(json_results)
def test_color_transform():
# test assertion for invalid type of magnitude
with pytest.raises(AssertionError):
transform = dict(type='ColorTransform', magnitude=None)
build_from_cfg(transform, PIPELINES)
# test assertion for invalid value of prob
with pytest.raises(AssertionError):
transform = dict(type='ColorTransform', magnitude=0.5, prob=100)
build_from_cfg(transform, PIPELINES)
# test assertion for invalid value of random_negative_prob
with pytest.raises(AssertionError):
transform = dict(type='ColorTransform',
magnitude=0.5,
random_negative_prob=100)
build_from_cfg(transform, PIPELINES)
# test case when magnitude=0, therefore no color transform
results = construct_toy_data_photometric()
transform = dict(type='ColorTransform', magnitude=0., prob=1.)
pipeline = build_from_cfg(transform, PIPELINES)
results = pipeline(results)
assert (results['img'] == results['ori_img']).all()
# test case when prob=0, therefore no color transform
results = construct_toy_data_photometric()
transform = dict(type='ColorTransform', magnitude=1., prob=0.)
pipeline = build_from_cfg(transform, PIPELINES)
results = pipeline(results)
assert (results['img'] == results['ori_img']).all()
# test case when magnitude=-1, therefore got gray img
results = construct_toy_data_photometric()
transform = dict(type='ColorTransform',
magnitude=-1.,
prob=1.,
random_negative_prob=0)
pipeline = build_from_cfg(transform, PIPELINES)
results = pipeline(results)
img_gray = mmcv.bgr2gray(results['ori_img'])
img_gray = np.stack([img_gray, img_gray, img_gray], axis=-1)
assert (results['img'] == img_gray).all()
# test case when magnitude=0.5
results = construct_toy_data_photometric()
transform = dict(type='ColorTransform',
magnitude=.5,
prob=1.,
random_negative_prob=0)
pipeline = build_from_cfg(transform, PIPELINES)
results = pipeline(results)
img_r = np.round(
np.clip((results['ori_img'] * 0.5 + img_gray * 0.5), 0,
255)).astype(results['ori_img'].dtype)
assert (results['img'] == img_r).all()
assert (results['img'] == results['img2']).all()
# test case when magnitude=0.3, random_negative_prob=1
results = construct_toy_data_photometric()
transform = dict(type='ColorTransform',
magnitude=.3,
prob=1.,
random_negative_prob=1.)
pipeline = build_from_cfg(transform, PIPELINES)
results = pipeline(results)
img_r = np.round(
np.clip((results['ori_img'] * 0.7 + img_gray * 0.3), 0,
255)).astype(results['ori_img'].dtype)
assert (results['img'] == img_r).all()
assert (results['img'] == results['img2']).all()
# -*- coding: utf-8 -* -
'''
MMCV读取并转换颜色空间
参考：https://mmcv.readthedocs.io/en/latest/image.html#color-space-conversion
'''
import mmcv
img = mmcv.imread("cluo.jpg")
img1 = mmcv.bgr2rgb(img)
img2 = mmcv.bgr2gray(img)
img3 = mmcv.bgr2gray(img)
