Welcome to DOFBOT Pro repository

Block Different Shape Sorting

Before starting this function, you need to close the large program and APP processes. If you need to restart the large program and APP later, start them from the terminal:

bash ~/dofbot_pro/APP_DOFBOT_PRO/start_app.sh

1. Function Description

After the program runs, place the same color blocks. After the camera captures the image, perform color recognition based on HSV values and recognize the shapes of the color blocks. Press the spacebar, the robotic arm will grasp the target shape color blocks and place them at the set position; after placement is complete, it returns to the initial color block recognition posture.

The wooden blocks used in this lesson are: 30x30x30mm, 30x30x60mm, and 30x30mm blocks.

2. Startup and Operation

(1). Startup Commands

Enter the following commands in the terminal to start:

xxxxxxxxxx
#Start camera:
ros2 launch orbbec_camera dabai_dcw2.launch.py
#Start underlying control:
ros2 run dofbot_pro_driver arm_driver
#Start inverse kinematics program:
ros2 run dofbot_pro_info kinemarics_dofbot
#Start robotic arm grasping program:
ros2 run dofbot_pro_driver grasp 
#Start color block shape recognition program:
ros2 run dofbot_pro_color shape_recognize

(2). Operation Process

After starting in the terminal, place color blocks of the same color. The camera captures the image. Press [r] or [R] to enter color selection mode, use the mouse to select a certain area of the color block, obtain the HSV values of this area, release the mouse to enter color recognition mode. After entering color recognition mode, if the current HSV values still cannot filter out other colors, you can fine-tune the HSV values through the dynamic parameter tuner. Enter the following command in the terminal to start the dynamic parameter tuner:

xxxxxxxxxx
ros2 run rqt_reconfigure rqt_reconfigure

You can modify the HSV values through sliders. When the left image (binary) only displays the uniquely recognized color, click the color image frame and press the spacebar, the robotic arm will lower the gripper to grasp color blocks of the set target shape. The default grasping target shape is square.

3. Program Flowchart

shape_recognize.py

4. Core Code Analysis

4.1. shape_recognize.py

Code path:

xxxxxxxxxx
/home/jetson/dofbot_pro_ws/src/dofbot_pro_color/dofbot_pro_color/shape_recognize.py

astra_common code library path:

xxxxxxxxxx
/home/jetson/dofbot_pro_ws/src/dofbot_pro_color/dofbot_pro_color/astra_common.py

Import necessary libraries

​x
import cv2
import rclpy
from rclpy.node import Node
import numpy as np
from message_filters import ApproximateTimeSynchronizer, Subscriber
from sensor_msgs.msg import Image
from std_msgs.msg import Float32, Bool
from cv_bridge import CvBridge
import cv2 as cv
​
import time
import math
import os
encoding = ['16UC1', '32FC1']
import tf_transformations as tf
import transforms3d as tfs
​
from dofbot_pro_color.astra_common import *
from dofbot_pro_interface.msg import *

Program parameter initialization, create publishers, subscribers, etc.

xxxxxxxxxx
def __init__(self):
    super().__init__('shape_detect')
    self.declare_param()
    self.target_servox=90
    self.window_name = "depth_image"
    self.target_servoy=45
    self.init_joints = [90.0, 120.0, 0.0, 0.0, 90.0, 90.0]
    #Create publisher for color block information
    self.pub_ColorInfo = self.create_publisher(AprilTagInfo, "PosInfo", 1)
    #Create publisher for robotic arm target angle
    self.pubPoint = self.create_publisher(ArmJoint, "TargetAngle", 1)
    #Create subscriber for gesture recognition results
    self.grasp_status_sub = self.create_subscription(Bool, 'grasp_done', self.GraspStatusCallback, 1)
    #Create two subscribers, subscribe to color image topic and depth image topic
    self.depth_image_sub  = Subscriber(self, Image, "/camera/color/image_raw", qos_profile=1)
    self.rgb_image_sub = Subscriber(self, Image, "/camera/depth/image_raw", qos_profile=1)
    self.TimeSynchronizer = ApproximateTimeSynchronizer([self.depth_image_sub, self.rgb_image_sub],queue_size=10,slop=0.5)
    self.TimeSynchronizer.registerCallback(self.TrackAndGrap)
    #Create bridges for converting color and depth image topic message data to image data
    self.rgb_bridge = CvBridge()
    self.depth_bridge = CvBridge()
​
    #color
    #Initialize region coordinates
    self.Roi_init = ()
    #Initialize HSV values
    self.hsv_range = ()
    #Initialize recognized color block information, here representing color block center x coordinate, center y coordinate and minimum enclosing circle radius r
    self.circle = (0, 0, 0)
    #Dynamic parameter adjustment flag, True means perform dynamic parameter adjustment
    self.dyn_update = True
    #Mouse selection flag
    self.select_flags = False
    self.gTracker_state = False
    self.windows_name = 'frame'
    self.Track_state = 'init'
    #Create color detection object
    self.color = color_detect()
    #Initialize row and column coordinates of region coordinates
    self.cols, self.rows = 0, 0
    #Initialize mouse selected xy coordinates
    self.Mouse_XY = (0, 0)
    #Default HSV threshold file path, this file stores the last saved HSV values
    self.hsv_text = "/home/jetson/dofbot_pro_ws/src/dofbot_pro_color/dofbot_pro_color/colorHSV.text"
    if os.path.exists(self.hsv_text): self.roi_hsv_range = read_HSV(self.hsv_text)
    #Flag for publishing AprilTag information, when True publish /PosInfo topic data
    self.pubPos_flag = False
    #Set target shape, default is cube, can choose rectangular prism and cylinder
    self.color.target_shape = "Square" # "Rectangle" ,"Cylinder"
    print("Target shape: ",self.color.target_shape)

Main image processing function TrackAndGrap

xxxxxxxxxx
def TrackAndGrap(self,color_frame,depth_frame):
    #rgb_image
    #Receive color image topic message, convert message data to image data
    rgb_image = self.rgb_bridge.imgmsg_to_cv2(color_frame,'bgr8')
    result_image = np.copy(rgb_image)
    #depth_image
    #Receive depth image topic message, convert message data to image data
    depth_image = self.depth_bridge.imgmsg_to_cv2(depth_frame, encoding[1])
    frame = cv.resize(depth_image, (640, 480))
    depth_image_info = frame.astype(np.float32)
    action = cv.waitKey(10) & 0xFF
    result_image = cv.resize(result_image, (640, 480))
    #Pass the obtained color image as parameter to process, and also pass keyboard event action
    result_frame, binary = self.process(result_image,action)
    #Determine if self.color.shape_cx, self.color.shape_cy are not 0, indicating there are color blocks meeting conditions
    if self.color.shape_cx!=0 and self.color.shape_cy!=0:
        pos = AprilTagInfo()
        pos.x = self.color.shape_cx
        pos.y = self.color.shape_cy
        print(self.color.shape_cx,self.color.shape_cy)
        pos.z = depth_image_info[self.color.shape_cy,self.color.shape_cx]/1000
        print("depth: ",pos.z)
        #Determine if self.pubPos_flag value is True and the color block's depth value is not 0, meeting both conditions means message can be published
        if self.pubPos_flag == True and pos.z!=0:
            self.pubPos_flag = False
            self.pub_ColorInfo.publish(pos)
    #Determine if binary image exists, if exists display color and binary images, otherwise only display color image
    if len(binary) != 0: cv.imshow(self.windows_name, ManyImgs(1, ([result_frame, binary])))
    else:
        cv.imshow(self.windows_name, result_image)

Image processing function self.process

xxxxxxxxxx
def process(self, rgb_img, action):
    rgb_img = cv.resize(rgb_img, (640, 480))
    binary = []
    #Determine key press event, when spacebar is pressed, change information publishing flag status, self.pubPos_flag True means information topic can be published
    if action == 32: self.pubPos_flag = True
    #Determine key press event, when i or I is pressed, change state, switch to recognition mode
    elif action == ord('i') or action == ord('I'): self.Track_state = "identify"
    #Determine key press event, when r or R is pressed, reset all parameters, enter color selection mode
    elif action == ord('r') or action == ord('R'): self.Reset()
    #Determine state value, if it's init, it means initial state value, at this time mouse can be used to select region
    if self.Track_state == 'init':
        cv.namedWindow(self.windows_name, cv.WINDOW_AUTOSIZE)
        #Select a region's color within the specified window
        cv.setMouseCallback(self.windows_name, self.onMouse, 0)
        #Determine color selection flag, true means color can be selected
        if self.select_flags == True:
            cv.line(rgb_img, self.cols, self.rows, (255, 0, 0), 2)
            cv.rectangle(rgb_img, self.cols, self.rows, (0, 255, 0), 2)
            #Determine if selected region exists
            if self.Roi_init[0] != self.Roi_init[2] and self.Roi_init[1] != self.Roi_init[3]:
                #Call Roi_hsv function in created color detection object self.color, returns processed color image and HSV values
                rgb_img, self.hsv_range = self.color.Roi_hsv(rgb_img, self.Roi_init)
                self.gTracker_state = True
                self.dyn_update = True
            else: self.Track_state = 'init'
    #Determine state value, if it's "identify", it means color recognition can be performed
    elif self.Track_state == "identify":
        #Determine if HSV threshold file exists, if exists read values inside and assign to hsv_range
        if os.path.exists(self.hsv_text): self.hsv_range = read_HSV(self.hsv_text)
        #If not exists, change state to init for color selection
        else: self.Track_state = 'init'
    if self.Track_state != 'init':
        #Determine length of self.hsv_range value, that is determine if this value exists, when length is not 0, enter color detection function
        if len(self.hsv_range) != 0:
            #Call ShapeRecognition function in created color detection object self.color, pass color image and self.hsv_range (hsv threshold), returns processed color image, binary image and information storing graphics matching hsv threshold, including center point coordinates and its minimum enclosing circle radius
            rgb_img, binary, self.circle = self.color.ShapeRecognition(rgb_img, self.hsv_range)
            #Determine dynamic parameter update flag, True means hsv_text file can be updated and values on parameter server can be modified
            if self.dyn_update == True:
                write_HSV(self.hsv_text, self.hsv_range)
                params = {'Hmin': self.hsv_range[0][0], 'Hmax': self.hsv_range[1][0],
                          'Smin': self.hsv_range[0][1], 'Smax': self.hsv_range[1][1],
                          'Vmin': self.hsv_range[0][2], 'Vmax': self.hsv_range[1][2]}
                self.dyn_client.update_configuration(params)
                self.dyn_update = False
    return rgb_img, binary

4.2. grasp.py

You can refer to section 4.2 [grasp.py] in tutorial [12. 3D Sorting and Grasping Course\1. AprilTag ID sorting].