Example #1
0
def draw_environment(path_to_follow, original_path, robot, costmap):
    """
    Draw obstacles, path and a robot
    :param path_to_follow: numpy array of (x, y, angle) of a path left to follow
    :param original_path: numpy array of (x, y, angle) of the original path that perhaps has been
                          followed up to some point.
    :param robot IRobot: the robot that will supply the draw function
    :param costmap Costmap: the costmap that will provide the obstacle data
    :return: numpy BGR rendered image
    """
    img = prepare_canvas(costmap.get_data().shape)

    draw_wide_path(img,
                   original_path,
                   robot_width=2 * inscribed_radius(robot.get_footprint()),
                   origin=costmap.get_origin(),
                   resolution=costmap.get_resolution(),
                   color=(240, 240, 240))

    if len(path_to_follow):
        draw_wide_path(img,
                       path_to_follow,
                       robot_width=2 * inscribed_radius(robot.get_footprint()),
                       origin=costmap.get_origin(),
                       resolution=costmap.get_resolution(),
                       color=(220, 220, 220))

    # Draw goal:
    goal_x, goal_y, goal_theta = original_path[-1]
    draw_robot(robot,
               img,
               goal_x,
               goal_y,
               goal_theta,
               color=(240, 240, 240),
               costmap=costmap,
               draw_steering_details=False)

    x, y, angle = robot.get_pose()
    draw_robot(robot, img, x, y, angle, color=(0, 100, 0), costmap=costmap)
    draw_world_map(img, costmap.get_data())

    # for pose in path_to_follow:
    #     x, y, angle = pose
    #     draw_robot(robot, img, x, y, angle, color=(100, 0, 0), costmap=costmap, alpha=0.1, draw_steering_details=False)

    return img
Example #2
0
def test_inscribed_radius():
    # square clockwise
    footprint = np.array([[-1, -1], [-1, 1], [1, 1], [1, -1]])
    assert abs(inscribed_radius(footprint) - 1.) < 1e-6

    # square anticlockwise
    footprint = np.array([[1, -1], [1, 1], [-1, 1], [-1, -1]])
    assert abs(inscribed_radius(footprint) - 1.) < 1e-6

    # square rotated
    footprint = np.array([[1, 0], [0, 1], [-1, 0], [0, -1]])
    assert abs(inscribed_radius(footprint) - np.sqrt(2) / 2) < 1e-6

    # triangle
    footprint = np.array([[1, -1], [-1, 1], [-1, -1]])
    assert abs(inscribed_radius(footprint) - 0.0) < 1e-6

    footprint = np.array([[2, -1], [-1, 1], [-1, -1]])
    assert abs(inscribed_radius(footprint) - 0.27735) < 1e-6

    # square with an appendix
    # The appendix has line segments 0.1 meters close if the distance is define by normals only
    footprint = np.array([[1, -1], [1, 1], [0.1, 1], [0.1, 1.1], [-0.1, 1.1],
                          [-0.1, 1.], [-1, 1], [-1, -1]])
    assert abs(inscribed_radius(footprint) - 1.) < 1e-6

    realistic_mule = np.array([
        # left side
        [1.13, 0.33],
        [0.85, 0.33],
        [0.85, 0.4675],
        [0.23, 0.4675],
        [0.23, 0.33],
        [-0.23, 0.33],
        # right side
        [-0.23, -0.33],
        [0.23, -0.33],
        [0.23, -0.4675],
        [0.85, -0.4675],
        [0.85, -0.33],
        [1.13, -0.33],
    ])
    assert abs(inscribed_radius(realistic_mule) - 0.23) < 1e-6

    beta_radius = inscribed_radius(_tricycle_footprint())
    assert np.abs(beta_radius - 0.3697) < 1e-3
def compute_cost_possibly_circumscribed_thresh(footprint, resolution,
                                               cost_scaling_factor):
    '''
    Compute costs of possibly circumscribed radius needed for some planners (e.g. sbpl)
    :param footprint: n x 2 numpy array with ROS-style footprint (x, y coordinates)
    :param resolution: costmap resolution
    :param cost_scaling_factor: 1 over the distance (in world units) at which costs
            beyond the inscribed radius decay by a factor of e.
    :return float: cost of possibly circumscribed radius
    '''
    distance = circumscribed_radius(footprint)
    inscribed_rad = inscribed_radius(footprint)
    distance_to_cost = _pixel_distance_to_cost(distance / resolution + 1,
                                               resolution, inscribed_rad,
                                               cost_scaling_factor)
    return max(0, distance_to_cost - 1)
def inflate_costmap(costmap, cost_scaling_factor, footprint):
    '''
    Inflate data with costs
    Costmap inflation is explained here:
    http://wiki.ros.org/costmap_2d#Inflation
    :param costmap CostMap2D: obstacle map
    :param cost_scaling_factor Float: how to scale the inflated costs
    :param footprint np.ndarray(N, 2)[Float]: footprint of the robot
    :return CostMap2D: inflated costs
    '''
    image_for_dist_transform = CostMap2D.LETHAL_OBSTACLE - costmap.get_data()
    distance = distance_transform(image_for_dist_transform)
    inscribed_rad = inscribed_radius(footprint)
    inflated_data = _pixel_distance_to_cost(distance, costmap.get_resolution(),
                                            inscribed_rad, cost_scaling_factor)

    return CostMap2D(data=inflated_data,
                     origin=costmap.get_origin(),
                     resolution=costmap.get_resolution())