perceive_world_old.py

#! /usr/bin/python

import math
import message_filters
import rospy

import cv2
from cv_bridge import CvBridge, CvBridgeError

from sensor_msgs.msg import Image, CameraInfo
from geometry_msgs.msg import Quaternion, Pose, PoseArray
from tf.transformations import quaternion_from_euler

from assistive_cane.utils import transform_pose
from assistive_cane.constants import CAMERA_TO_BOX_TOP, CAMERA_TO_CANE_TOP
from assistive_cane.world_visualizer import WorldVisualizer

bridge = CvBridge()
visualize_detected_objects = True

"""
Code Implementation Overview

Day 1: TODO 1, TODO 2, TODO 3

Day 2: TODO 4, TODO 5

Day 3: None

Day 4: None

Day 5: Demo!
"""

def pixel2world(camera_info, image_x, image_y, orie_x, orie_y, depth):
    """
    Returns pixel location and yaw in world space using camera and depth info

    Parameters
    ----------
    camera_info: CameraInfo
        contains meta information for overhead cameras. 
        For more info see: http://docs.ros.org/en/noetic/api/sensor_msgs/html/msg/CameraInfo.html
    image_x: int
        x center coordinate of marker
    image_y: int
        y center coordinate of marker
    orie_x: int
        x orientation of marker
    orie_y: int
        y oreintatin of marker

    depth: int
        depth of pixel (x,y)

    Returns
    -------
    tuple
        a tuple of pixel location in world space and yaw (world_x, world_y, world_z, yaw)
    """

    # Get focal lengths in pixel units (fx, fy) and principal 
    # and points (cx, cy)
    # Reference: https://docs.opencv.org/2.4/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html
    fx = camera_info.K[0]
    fy = camera_info.K[4]
    cx = camera_info.K[2]
    cy = camera_info.K[5]

    """
    [Day 2] TODO 5: Convert from pixel to world space and calculate yaw
        - Use the reference above to understand the math needed to convert
    """
    ####### Insert Code Here #######

    # Convert to world space
    world_x = (image_x - cx) * depth / fx
    world_y = (image_y - cy) * depth / fy
    world_z = depth

    # Calculate yaw, make sure to use radians!
    yaw = math.atan2(orie_y, orie_x)
    
    ################################

    return (world_x, world_y, world_z, yaw)


def image_callback(rgb_msg, camera_info):
    """
    Goes through input image and locates cane and obstacle markers

    Parameters
    ----------
    rgb_msg: Image
        ROS message that contains image
        For more info see: http://docs.ros.org/en/noetic/api/sensor_msgs/html/msg/Image.html
    camera_info: CameraInfo
        ROS message that contains meta information for overhead camera. 
        For more info see: http://docs.ros.org/en/noetic/api/sensor_msgs/html/msg/CameraInfo.html

    Returns
    -------
    None
        Publishes a list of ROS Pose messages describing located obstacles 
        and cane using boxes_state_publisher and cane_state_publisher, respectively
    """
    print("In image callback")
    
    try:
        # Convert your ROS Image message to OpenCV2
        cv2_img = bridge.imgmsg_to_cv2(rgb_msg, "bgr8")
    except e:
        print(e)
    else:
        # Using this ArUCo tutorial: https://pyimagesearch.com/2020/12/21/detecting-aruco-markers-with-opencv-and-python/

        # NOTE: Verify on end of day 1 image is displayed
        image = cv2_img

        """
        [Day 2] TODO 4: Detect markers
            Use the tutorial above to
            - Get an aruco dictionary
            - Get aruco parameters
            - Perform detection
        """
        ####### Insert Code Here ####### Day 2
        # Get dictionary of ArUco markers being used
        arucoDict = cv2.aruco.Dictionary_get(cv2.aruco.DICT_4X4_50)
        # Define ArUco detection parameters
        arucoParams = cv2.aruco.DetectorParameters_create()
        # Perform ArUco marker detection
        (corners, ids, rejected) = cv2.aruco.detectMarkers(image, arucoDict, parameters=arucoParams)

        ################################
        if corners == None:
            cv2.imshow("Image". cv2_img)
            cv2.waitKey(100)
            return None
            
        box_poses = []

        is_cane_detected = False
        cane_pose = Pose()

        # If markers detected...
        if len(corners) > 0:
            print("Detected arUco")

            # flatten the ArUco IDs list
            ids = ids.flatten()

            # loop over the detected ArUCo corners
            for (markerCorner, markerID) in zip(corners, ids):
                
                # extract the marker corners (which are always returned in
                # top-left, top-right, bottom-right, and bottom-left order)
                corners = markerCorner.reshape((4, 2))
                (topLeft, topRight, bottomRight, bottomLeft) = corners.astype(int)

                # compute the center (x, y)-coordinates of the ArUco marker
                centerX = (topLeft[0] + bottomRight[0]) // 2
                centerY = (topLeft[1] + bottomRight[1]) // 2
                
                # compute the top-middle (x, y)-coordinates of the ArUco marker
                forwardX = (topLeft[0] + topRight[0]) // 2
                forwardY = (topLeft[1] + topRight[1]) // 2

                ### Visualization ####

                # draw the bounding box of the ArUCo detection
                cv2.line(image, topLeft, topRight, (0, 255, 0), 2)
                cv2.line(image, topRight, bottomRight, (0, 255, 0), 2)
                cv2.line(image, bottomRight, bottomLeft, (0, 255, 0), 2)
                cv2.line(image, bottomLeft, topLeft, (0, 255, 0), 2)
                
                # visualize center of ArUco marker
                cv2.circle(image, (centerX, centerY), 4, (0, 0, 255), -1)
                
                # visualize top of ArUco marker
                cv2.circle(image, (forwardX, forwardY), 4, (0, 0, 255), -1)
                
                # draw the ArUco marker ID on the image
                cv2.putText(image, str(markerID), (topLeft[0], topLeft[1] - 15), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

                print("[INFO] ArUco marker ID: {}".format(markerID))   

                # Get x and y reference vectors
                orie_x = int((topLeft[0]+topRight[0])/2.0 - (bottomLeft[0] + bottomRight[0])/2.0)
                orie_y = int((topLeft[1]+topRight[1])/2.0 - (bottomLeft[1] + bottomRight[1])/2.0)

                # If marker is for obstacle (box)
                if markerID == 1:

                    # Get world coordinates of box
                    world_x, world_y, world_z, yaw = pixel2world(camera_info, centerX, centerY, orie_x, orie_y, CAMERA_TO_BOX_TOP)

                    # Convert to quaternion
                    q = quaternion_from_euler(0, 0, yaw)

                    # Create pose messsage
                    pose = Pose()
                    pose.position.x = world_x
                    pose.position.y = world_y
                    pose.position.z = world_z
                    pose.orientation = Quaternion(q[0], q[1], q[2], q[3])

                    pose = transform_pose(pose, "camera_color_optical_frame", "map")
                    box_poses.append(pose)

                # If marker is for cane
                elif markerID == 0:
                    
                    is_cane_detected = True

                    # Get world coordinates of cane
                    world_x, world_y, world_z, yaw = pixel2world(camera_info, centerX, centerY, orie_x, orie_y, CAMERA_TO_CANE_TOP)

                    q = quaternion_from_euler(0, 0, yaw)

                    cane_pose.position.x = world_x
                    cane_pose.position.y = world_y
                    cane_pose.position.z = world_z
                    cane_pose.orientation = Quaternion(q[0], q[1], q[2], q[3])

                    cane_pose = transform_pose(cane_pose, "camera_color_optical_frame", "map")
            
            # show the output image
            cv2.imshow("Image", image)
            cv2.waitKey(100)

        
            if is_cane_detected:
                cane_state_publisher.publish(cane_pose)
                if visualize_detected_objects:
                    world_visualizer.visualize_cane(cane_pose)

            box_pose_array = PoseArray()
            box_pose_array.poses = box_poses

            boxes_state_publisher.publish(box_pose_array)    

            if len(box_poses) != 0 and visualize_detected_objects:
                world_visualizer.visualize_boxes(box_poses)
        else:
            cv2.imshow("Image", cv2_img)
            cv2.waitKey(100)

def main():
    
    """
    [Day 1] TODO 1: Write code for the following
        - Initialize a ROS node called "perceive_world"
        - Define variables for the following topics:
            - "/camera/color/image_raw"
            - "/camera/color/camera_info"

    """
    ####### Insert Code Here #######

    # Initialize node
    rospy.init_node('perceive_world', anonymous=True)
    
    # Define image topics
    image_topic = "/camera/color/image_raw"
    camera_info_topic = "/camera/color/camera_info"

    ################################
    
    # Set up your subscriber and define its callback
    # rospy.Subscriber(image_topic, Image, image_callback)
    # Spin until ctrl + c
    image_sub = message_filters.Subscriber(image_topic, Image)
    info_sub = message_filters.Subscriber(camera_info_topic, CameraInfo)
    ts = message_filters.ApproximateTimeSynchronizer([image_sub, info_sub], 10, 0.2)
    ts.registerCallback(image_callback)

    global world_visualizer
    world_visualizer = WorldVisualizer()

    global boxes_state_publisher
    """
    [Day 1] TODO 2: Create boxes state publisher
        - Create a publisher with the following parameters
            - Topic: "boxes_state"
            - Message type: PoseArray
            - Queue size: 10

    """
    ####### Insert Code Here #######

    boxes_state_publisher = rospy.Publisher("boxes_state", PoseArray, queue_size=10)

    ################################

    global cane_state_publisher
    """
    [Day 1] TODO 3: Create cane state publisher
        - Create a publisher with the following parameters
            - Topic: "cane_state"
            - Message type: Pose
            - Queue size: 10

    """
    ####### Insert Code Here #######
    
    cane_state_publisher = rospy.Publisher("cane_state", Pose, queue_size=10)
    
    ################################

    rospy.spin()

if __name__ == '__main__':
    main()