get world coordinates from image coordinates

edit

hi everyone, I now maybe someone asked this question before but I can't found my answer in this forum.
I have a calibrated virtual single camera in gazebo8 and ROS kinetic, after applying some opencv filters, now an object was detected in x = 79.0, y = 23.0, w = 28.0, h = 127.0 coordinates of image. and was published this info into /robot0/object topic.
I'm using image_geometry package to get position of that object in gazebo simulated world, also I know position of that object in this case x = 245 , y = 269 ,I'm using gmapping package and size of map has set to 500 , 500.
this is my code that get object info from /robot0/object topic and try to guess object position in gazebo simulated world :

def get_object(array):
    global img_geo,robot_odometry
    u = array.data[0] + (array.data[2] // 2)
    v = array.data[1] + array.data[3]
    time = rospy.Time(0)
    listener = tf.listener.TransformListener()
    camera_point = img_geo.projectPixelTo3dRay((img_geo.rectifyPoint((u, v))))
    point_msg.pose.position.x = camera_point[0] + robot_odometry.pose.pose.position.x
    point_msg.pose.position.y = camera_point[1] + robot_odometry.pose.pose.position.y
    point_msg.pose.position.z = 0
    point_msg.pose.orientation.x = 0
    point_msg.pose.orientation.y = 0
    point_msg.pose.orientation.z = 0
    point_msg.pose.orientation.w = 1
    point_msg.header.frame_id = img_geo.tfFrame()
    point_msg.header.stamp = time
    try:
        listener.waitForTransform(img_geo.tfFrame(), 'map', time, rospy.Duration(1))
        tf_point = listener.transformPose('map', point_msg)
        print(convert_from_robot_to_map(tf_point.pose.position.y, tf_point.pose.position.x))
    except Exception:
        pass

def convert_from_robot_to_map(robot_y, robot_x):
    global map_info
    map_x = (robot_x - map_info.info.origin.position.x) / map_info.info.resolution
    map_y = (robot_y - map_info.info.origin.position.y) / map_info.info.resolution
    return map_y, map_x

but unfortunately printed result isn't true. printed info (print(convert_from_robot_to_map(...))) is y = 250 , x = 249 that is almost robot position in map( robot position in this case is x = 250 , y = 250), and camera_point object from projecting pixel to 3d ray in this case is (-0.23228537548412242, 0.0925648488771315, 0.9682330572174006) . and tf_point object values is(-2.3449200226706126, 0.910935376894018).
I also passed my camera info to img_geo.fromCameraInfo().
What is wrong in this code?? what should I do to get true coordinates of object in gazebo world???