def next(self):
# just grab the next random frame
frame = self.source[random.choice(self.frame_ids)]
# frame.show_image_with_labels()
# get a random crop object
crop_objs = filter(lambda x: not self.obj_types or x.obj_type in self.obj_types,frame.get_objects())
# print 'Num crop objs in sample: {0}'.format(len(crop_objs))
crop = random.choice(crop_objs)
# print 'crop_box: ' + str(crop.box)
# frame.show_image_with_labels()
# 1) Randomly perturb crop box (scale and translation)
transformed_box = RandomPerturber.perturb_crop_box(crop.box,{})
# 2) Take crop, todo, change to using center crop to preserve aspect ratio
# check if the affine is identity within some toleranc, then don't bother applying
affine = Affine()
scalex = float(self.crop_size[0])/transformed_box.edges()[0]
scaley = float(self.crop_size[1])/transformed_box.edges()[1]
affine.append(Affine.translation(-transformed_box.xy_min()))
affine.append(Affine.scaling((scalex,scaley)))
transformed_image = affine.apply_to_image(frame.image,self.crop_size)
# transformed_image.visualize(title='transformed image')
# 3) Randomly perturb cropped image (rotation only)
chw_image = transformed_image.to_order_and_class(Ordering.CHW,ValueClass.FLOAT01)
# chw_image.visualize(title='chw_image')
sample = AutoEncoderSample([torch.Tensor(chw_image.get_data().astype(float))],
[torch.Tensor(chw_image.get_data().astype(float))])
return sample
def crop_image_resize(ptimage, crop_box, output_resolution):
affine = Affine()
scale = min(
float(output_resolution[0]) / crop_box.edges()[0],
float(output_resolution[1]) / crop_box.edges()[1])
crop_offset = crop_box.xy_min() + 0.5 * (
np.array(crop_box.edges()) - np.array(output_resolution) / scale)
affine.append(Affine.translation(-crop_offset))
affine.append(Affine.scaling([scale, scale]))
# print affine.transform
return affine
def resize_image_center_crop(ptimage, output_resolution):
affine = Affine()
# calculate scale and crop offsets
wh = ptimage.get_wh()
scale = min(
float(output_resolution[0]) / wh[0],
float(output_resolution[1]) / wh[1])
offset = 0.5 * (scale * np.array(wh) - np.array(output_resolution))
affine.append(Affine.scaling([scale, scale]))
affine.append(Affine.translation(-offset))
return affine
def generate_random_affine(center,edges,params):
affine = Affine()
# translate to center coordinates
affine.append(Affine.translation(-center))
translate = get_deep(params,'translation_range',[-0.1,0.1])
tx = random.random()*(translate[1]-translate[0])+translate[0]
ty = random.random()*(translate[1]-translate[0])+translate[0]
translation_range = np.array([tx,ty])*(-edges)
affine.append(Affine.translation(translation_range))
scale = get_deep(params,'scaling_range',[0.9,1.4])
scale = random.random()*(scale[1]-scale[0])+scale[0]
scaling_range = np.array([scale,scale])
affine.append(Affine.scaling(scaling_range))
# translate back
affine.append(Affine.translation(center))
return affine