def add_background(self, back_img, patch_location, mode='resize'):
'''
add random background to a specific path in the image
parameters:
back_img: integer image
patch_location: [x_left, y_lower, x_right, y_upper], a location of the patch
mode: trun, keep, resize
'''
if isstring(back_img):
back_img = load_image(back_img, mode='numpy', debug=self.debug)
if self.debug:
assert isuintimage(
back_img), 'the input background image is not correct'
assert patch_location[0] >= 0 and patch_location[
1] >= 0, 'the input patch location is not correct'
assert patch_location[2] <= self.width and patch_location[
3] <= self.height, 'the input patch location is not correct'
assert patch_location[2] >= patch_location[0] and patch_location[
3] >= patch_location[
1], 'the input patch location is not correct'
assert mode == 'trim' or mode == 'resize', 'the input mode is not correct'
if mode == 'resize':
back_img = imresize(back_img,
(patch_location[3] - patch_location[1],
patch_location[2] - patch_location[0], 3))
self.img[patch_location[1]:patch_location[3],
patch_location[0]:patch_location[2], :] = np.copy(
back_img)
def generate_video_from_list(image_list, save_path, framerate=30, downsample=1, warning=True, debug=True):
'''
create video from a list of images with a framerate
note that: the height and widht of the images should be a multiple of 2
parameters:
image_list: a list of image path
save_path: the path to save the video file
framerate: fps
'''
if debug:
assert islistofstring(image_list), 'the input is not correct'
assert ispositiveinteger(framerate), 'the framerate is a positive integer'
mkdir_if_missing(save_path)
inputdict = {'-r': str(framerate)}
outputdict = {'-r': str(framerate), '-crf': '18', '-vcodec': 'libx264', '-profile:V': 'high', '-pix_fmt': 'yuv420p'}
video_writer = FFmpegWriter(save_path, inputdict=inputdict, outputdict=outputdict)
count = 1
num_images = len(image_list)
for image_path in image_list:
print('processing frame %d/%d' % (count, num_images))
image = load_image(image_path, resize_factor=downsample, warning=warning, debug=debug)
# make sure the height and width are multiple of 2
height, width = image.shape[0], image.shape[1]
if not (height % 2 == 0 and width % 2 == 0):
height += height % 2
width += width % 2
image = image_resize(image, target_size=[height, width], warning=warning, debug=debug)
video_writer.writeFrame(image)
count += 1
video_writer.close()
def test_triangulate_multiple_views():
calibration_file = '../calibration.txt'
camera_cluster = load_camera_cluster(calibration_file, warning=False)
pts1_file = '../20180514--handsy--400015--0800.pts'
pts_array1 = anno_parser(pts1_file)
pts2_file = '../20180514--handsy--400053--0800.pts'
pts_array2 = anno_parser(pts2_file)
pts3_file = '../20180514--handsy--400025--0800.pts'
pts_array3 = anno_parser(pts3_file)
pts4_file = '../20180514--handsy--400027--0800.pts'
pts_array4 = anno_parser(pts4_file)
pts_array = np.zeros((4, 3, 21), dtype='float32')
# print(pts_array)
pts_array[0, :, :] = pts_array1
# print(pts_array)
pts_array[1, :, :] = pts_array2
# print(pts_array)
pts_array[2, :, :] = pts_array3
pts_array[3, :, :] = pts_array4
# pts_array = np.vstack((pts_array1, pts_array2, pts_array3, pts_array4))
projection1 = camera_cluster['400015'].get_projection_matrix()
projection2 = camera_cluster['400053'].get_projection_matrix()
projection3 = camera_cluster['400025'].get_projection_matrix()
projection4 = camera_cluster['400027'].get_projection_matrix()
projection = np.zeros((4, 3, 4), dtype='float32')
projection[0, :, :] = projection1
projection[1, :, :] = projection2
projection[2, :, :] = projection3
projection[3, :, :] = projection4
# projection = np.vstack((projection1, projection2, projection3, projection4))
# print(pts_array.shape)
# pts_3d, pts_reproj = triangulate_multiple_views(pts_array, projection, scaling_factor=1000)
pts_3d, pts_reproj = triangulate_multiple_views(pts_array,
projection,
scaling_factor=1)
img_path = '../20180514--handsy--400015--0800.png'
# img_path = '../20180514--handsy--400053--0800.png'
# img_path = '../20180514--handsy--400025--0800.png'
# img_path = '../20180514--handsy--400027--0800.png'
img = load_image(img_path)
debug = False
fig, ax = visualize_image_with_pts(img,
pts_array[0, :, :],
label=True,
debug=debug,
closefig=False,
vis=True)
fig, ax = visualize_image_with_pts(img,
pts_reproj[0, :, :],
label=True,
debug=debug,
closefig=False,
vis=True)
# visualize_pts_array(pts_reproj[0, :], fig=fig, ax=ax, color_index=1, pts_size=20, label=True, vis=True, debug=debug)
print('\n\nDONE! SUCCESSFUL!!\n')
def __init__(self, img, debug=True):
'''
generate synthetic data on a given image, the image should be an numpy array
'''
if isstring(img):
img = load_image(img, mode='numpy', debug=debug)
if debug:
assert isuintimage(img), 'the input image is not correct'
self.debug = debug
self.img = img
self.width = img.shape[1]
self.height = img.shape[0]
self.__original = np.copy(img)
def test_triangulate_two_views():
calibration_file = '../calibration.txt'
camera_cluster = load_camera_cluster(calibration_file, warning=False)
pts1_file = '../20180514--handsy--400015--0800.pts'
pts_array1 = anno_parser(pts1_file)
pts2_file = '../20180514--handsy--400053--0800.pts'
pts_array2 = anno_parser(pts2_file)
projection1 = camera_cluster['400015'].get_projection_matrix()
projection2 = camera_cluster['400053'].get_projection_matrix()
# pts_array1 = np.array([[54], [183], [1]]).astype('float32')
# projection1 = np.array([[1520.4, 0, 302.3, 0], [0, 1525.9, 246.9, 0], [0, 0, 1, 0]])
# pts_array2 = np.array([[54], [199], [1]])
# projection2 = np.array([[1520.9, -27.4, 298.7, 5.3], [-49.3, 1410.6, 630.2, -1488.8], [0, -0.3, 1, 0.2]])
pts_3d, pts1_reproj, pts2_reproj = triangulate_two_views(pts_array1,
pts_array2,
projection1,
projection2,
scaling_factor=1)
img_path = '../20180514--handsy--400015--0800.png'
img = load_image(img_path)
debug = False
fig, ax = visualize_image_with_pts(img,
pts_array1,
label=True,
debug=debug,
closefig=False,
vis=True)
visualize_image_with_pts(img,
pts1_reproj,
label=True,
debug=debug,
closefig=False,
vis=True)
# visualize_pts_array(pts1_reproj, fig=fig, ax=ax, color_index=1, pts_size=20, label=True, vis=True, debug=debug)
print('\n\nDONE! SUCCESSFUL!!\n')
save_dir = os.path.join(data_dir, 'centered122_flow/ABOUT/train/ABOUT_00001'); mkdir_if_missing(save_dir)
image_list, num_images = load_list_from_folder(images_dir)
print('number of images loaded is %d' % num_images)
test = cv2.imread(image_list[0])
# print(test.dtype)
test = cv2.cvtColor(test, cv2.COLOR_BGR2GRAY)
# print(test.dtype)
# print(test)
# zxc
frame1 = load_image(image_list[0])
prvs = rgb2gray(frame1) # uint8
# print(prvs)
hsv = np.zeros_like(frame1)
hsv[..., 1] = 255
image_index = 1
while(1):
image_path = image_list[image_index]
_, filename, _ = fileparts(image_path)
frame2 = load_image(image_path)
next = rgb2gray(frame2)
# save_image(next, '/home/xinshuo/aa.jpg')
##--------------------------------- Testing ----------------------------------##
image_list, num_list = load_list_from_folder(images_dir,
ext_filter=['.png', '.jpg'],
depth=2)
print_log('testing results on %d images' % num_list, log=log_file)
count = 1
timer = Timer()
timer.tic()
for image_file_tmp in image_list:
parent_dir, filename, _ = fileparts(image_file_tmp)
video_dir = parent_dir.split('/')[-1]
# print(video_dir)
# zxc
image = load_image(image_file_tmp)
results = model.detect([image]) # inference, results is a dictionary
if len(results) == 0:
count += 1
print_log('Mask-RCNN demo: testing %d/%d, no detected results!!!!!' %
(count, num_list),
log=log_file)
continue
# visualize and save results
r = results[0] # results from the first image
num_instances = r['masks'].shape[-1]
bboxes_tmp = r['rois'] # y1, x1, y2, x2 format
bboxes_tmp[:, [0, 1]] = bboxes_tmp[:, [1, 0]]
bboxes_tmp[:, [2, 3]] = bboxes_tmp[:, [3, 2]] # x1, y1, x2, y2 format