def recognize():
" Compute mesh from images (pure mode aware)"
view_stop()
calibration_data = settings.load_data(CalibrationData())
r = settings.get_laser_range()
slices, colors = cloudify(calibration_data, settings.WORKDIR, r, range(360), method=settings.SEGMENTATION_METHOD)
meshify(calibration_data, slices, colors=colors, cylinder=settings.ROI).save("model.ply")
gui.clear()
def scan_object():
""" Scan object """
calibration_data = settings.load_data(CalibrationData())
r = settings.get_laser_range()
cloudifier = iter_cloudify(calibration_data, settings.WORKDIR, r, range(360), method=settings.SEGMENTATION_METHOD)
iterator = partial(next, cloudifier)
capture(on_step=iterator, display=False, ftw=settings.SYNC_FRAME_FAST)
slices, colors = iterator()
meshify(calibration_data, slices, colors=colors).save("model.ply")
gui.clear()
def calibration(calibration_data, images):
obj_points = []
img_points = []
found_nr = 0
failed_serie = 0
term = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 100, 0.001)
flags = cv2.CALIB_CB_FAST_CHECK
pattern_points = settings.get_pattern_points()
temp_calibration_data = defaultdict(lambda: {})
for idx, fn in enumerate(images):
gui.progress('Webcam calibration %s (%d found)... ' % (fn, found_nr), idx, len(images))
img, hsv = imtools.imread(fn, format="full")
grey = hsv[:,:,2]
if img is None:
print("Failed to load", fn)
continue
w, h = img.shape[:2]
found, corners = chess_detect(grey, flags)
if not found:
if found_nr > 20 and failed_serie > 6:
break
failed_serie += 1
continue
failed_serie = 0
found_nr += 1
cv2.cornerSubPix(grey, corners, (11, 11), (-1, -1), term)
temp_calibration_data[fn]['chess_corners'] = corners
img_points.append(corners)
obj_points.append(pattern_points)
# compute mask coordinates
p1 = corners[0][0]
p2 = corners[settings.PATTERN_MATRIX_SIZE[0] - 1][0]
p3 = corners[settings.PATTERN_MATRIX_SIZE[0] * (settings.PATTERN_MATRIX_SIZE[1] - 1)][0]
p4 = corners[settings.PATTERN_MATRIX_SIZE[0] * settings.PATTERN_MATRIX_SIZE[1] - 1][0]
temp_calibration_data[fn]['chess_contour'] = np.array([p1, p2, p4, p3], dtype='int32')
if idx%settings.ui_base_i == 0:
chess_draw(img, found, corners)
gui.display(img, 'chess', resize=True)
if settings.skip_calibration:
print("\nskipping camera calibration...")
try:
settings.load_data(calibration_data)
except Exception:
# Data for Logitech C270
calibration_data.camera_matrix = np.array(([1430.0, 0.0, 480.0], [0.0, 1430.0, 620.0], [0.0, 0.0, 1.0]))
calibration_data.distortion_vector = np.array((0.0, 0.0, 0.0, 0.0, 0.0))
return temp_calibration_data
print("\nComputing camera calibration...")
if not obj_points:
raise ValueError("Unable to detect pattern on screen :(")
rms, camera_matrix, dist_coefs, rvecs, tvecs = cv2.calibrateCamera(obj_points, img_points, grey.shape, None, None)
if rms:
error = 0
# Compute calibration error
for i in range(len(obj_points)):
imgpoints2, _ = cv2.projectPoints(obj_points[i], rvecs[i], tvecs[i], camera_matrix, dist_coefs)
error += abs(
cv2.norm(img_points[i])
- cv2.norm(imgpoints2)
)
error /= len(obj_points)
print("Camera calibration error = %.4fmm"%error)
calibration_data.camera_matrix = camera_matrix
calibration_data.distortion_vector = dist_coefs.ravel()
settings.save_data(calibration_data)
return temp_calibration_data