1、首先进入python3的环境并安装相关依赖包
sudo apt-get install python-catkin-tools python3-dev python3-catkin-pkg-modules python3-numpy python3-yaml ros-melodic-cv-bridge
2、创建一个工作空间用于存放待编译的 cv_bridge 文件
mkdir -p catkin_workspace/src
3、指示carkin设置cmake变量
cd catkin_workspace #注意，该处需要根据你原本自带的系统python版本设定。16.04是python3.5,18.04是python3.6 catkin config -DPYTHON_EXECUTABLE=/usr/bin/python3 -DPYTHON_INCLUDE_DIR=/usr/include/python3.6m -DPYTHON_LIBRARY=/usr/lib/x86_64-linux-gnu/libpython3.6m.so catkin config -DPYTHON_EXECUTABLE=/usr/bin/python3 -DPYTHON_INCLUDE_DIR=/usr/include/python3.6m -DPYTHON_LIBRARY=/usr/lib/aarch64-linux-gnu/libpython3.6m.so apt-cache show ros-kinetic-cv-bridge | grep Version sudo apt-get install ros-melodic-cv-bridge
4、Instruct catkin to install built packages into install place。这一步不成功也没关系，可不用。
# Instruct catkin to install built packages into install place. It is $CATKIN_WORKSPACE/install folder catkin config --install
5、在catkin_workspace工作空间中克隆 cv_bridge src
git clone https://github.com/ros-perception/vision_opencv.git src/vision_opencv
6、Find version of cv_bridge in your repository
apt-cache show ros-melodic-cv-bridge | grep Version
7、Checkout right version in git repo. In our case it is 1.12.8
#该步骤我是省去的 #必须要编译否则cv_bridge的CMakeLists默认版本为python37 cd src/vision_opencv/ git checkout 1.13.0 cd ../../
8、开始编译
#有可能会有warning，不用管，只要不报错就行 catkin build 或者 catkin build cv_bridge
9、进入python3环境(virtualenv)之后，先进入到catkin_workspace工作目录下，运行下面的source，然后再到相关的节点工作空间（如catkin_ws），就可以启动那些使用到cv_bridge库的相关节点了：
# 打开虚拟环境env_py3和进入catkin_workspace空间进行source。 conda activate pytorch cd catkin_workspace/ # 这里 --extend 参数的作用是让这次的路径配置不影响之前配置好的路径，否则这一次source会覆盖掉之前配置的路径。 source install/setup.bash --extend #每次都必须先激活然后才能使python3调用cv_bridge cd .. # 进入另一个工作空间，该空间含有需要启动的python脚本节点。 cd catkin_ws/src/beginner_tutorials/scripts/ python ImgSub.py二、报错boost_python3：
编译过程中，如果出现下面的报错(目前好像vision_opencv导出来的boost是boost_python37)：
Could not find the following Boost libraries:
boost_python3
No Boost libraries were found. You may need to set BOOST_LIBRARYDIR to the
directory containing Boost libraries or BOOST_ROOT to the location of
Boost.
**原始行：**
find_package(Boost REQUIRED python3)
更改成：
find_package(Boost REQUIRED python-py35)三、报错libgcc_s.so.1
sudo apt install libgcc1-dbg四、丢包、延迟等问题
主要原因，可能是因为用的是WiFi无线网络，带宽有限，视频流数据庞大，所以经常会丢包。获取不到话题。
所以没办法，我只能订阅压缩过的话题——
#!/usr/bin/python
# encoding:utf-8
# 导入消息类型，压缩的和未压缩的
import time
from sensor_msgs.msg import CompressedImage
from cv_bridge import CvBridge, CvBridgeError
import rospy
class Detect(object):
def __init__(self):
self.start_time = time.time()
# 订阅网络压缩话题，并且，消息类型换成压缩的
self.image_sub = rospy.Subscriber("/kinect2/hd/image_color_rect/compressed", CompressedImage, self.rgb_callback,
queue_size=1)
self.depth_sub = rospy.Subscriber("/kinect2/hd/image_depth_rect/compressed", CompressedImage,
self.depth_callback, queue_size=1)
self.bridge = CvBridge()
def rgb_callback(self, image):
try:
# 尝试转换压缩图片信息到CV2可以直接用的格式。后面加不加bgr8好像都行。
self.rgb_image = self.bridge.compressed_imgmsg_to_cv2(image, "bgr8")
# 用上面的和下面的都行，都是会转成rgb的图，但是对于第二个深度图就不合适了，只能用上面的
# np_arr = np.fromstring(image.data,np.uint8)
# opencv3 is this ,and opencv2 is cv2.CV_LOAD_IMAGE_COLOR
# self.rgb_image = cv2.imdecode(np_arr,cv2.IMREAD_COLOR)
except CvBridgeError as e:
print(e)
rospy.loginfo('convert rgb image error')
def depth_callback(self, depth):
try:
print("depth_start:")
self.depth_image = self.bridge.compressed_imgmsg_to_cv2(depth)
print(self.depth_image.shape)
except CvBridgeError as e:
print(e)
rospy.loginfo('convert depth image image error')
