robertgcniu / uav-gesture-control_python Goto Github PK

Python 76.59% Shell 0.39% CMake 2.28% Makefile 20.74%

uav-gesture-control_python's Introduction

2018/12/20 Update

The EDU version of Tello has been released. The new function of Tello EDU is AP command. We can send this command to Tello to transfter the it as a station

AP YourWiFiName YourWiFiPW

Then you can log in your router to see the corresponding IP address of UAVs. My UAVs' IPs are '192.168.1.239' and '192.168.1.163'. Then We can send the command message to Tello. A demo for taking off is shown here. The code is here

UAV-Gesture-Control

This project is made by Chinese Univeristy of Hong Kong, Shenzhen(CUHKSZ) and Stanford.

The goal is to control the UAV Tello by hand gesture or body pose.

Simple Control

You can use the files in Simple-Control. The message is sent as string to Tello and the response will feed back to PC. The new SDK is available now. I am not sure if all the commands can still work. You can easily modify the corresponding command if anything different in new SDK.

Tellopy Control

The Tellopy can be installed by pip. But When I intall 'av' by pip, the error occured and I can't fix it. The easy method to solve is to use Anaconda by running

$conda install av -c conda-forge

The interesting thing is that 'av' can be easily installed without error when I use Python 2.7.12 by running

$pip install av

Openpose

Please follow the Installation Manual to install Openpose. Ubuntu version is recommended. All of codes are only checked in Ubuntu 16.04 system. Please make sure all of the required packages have been installed. It will cost almost 30 minutes.

If you have error like 'CMake Error at /usr/share/cmake-3.5/Modules/ExternalProject.cmake:1915 (message)' when running Cmake GUI, please run the following code

$cd 3rdparty
$git clone https://github.com/CMU-Perceptual-Computing-Lab/caffe.git

Now, run CMake GUI to Configure and Generate.

Check OpenPose was properly installed by running it on the default images, video, or webcam: Quick Start.

Pose Detection by the Camera of Tello

The python codes are updated and can be run directly if you connect your Tello with PC. Please copy the files '/python_tutorial' to your computer path 'openpose/build/examples/python_tutorial'

The following video is the result of video_effect.py.

Use kNN to classify the pose

The training data sets are collected by 'mat' file. All codes and data sets are here

Speech control

I try to use Xun Fei Yun to realize the off-line speech control. However, it only provides C++ API and it's not easy for me to transfer as Python version.

Speech_Recognition package is a strong tool for speech recognition. However, there are so many bugs when I try to install it on Ubuntu 16.04. I firstly use PC and error told me that microphone can't be found. I used a USB microphone. It can't be recognize by Pyaudio. After a long-time struggle, I decide to use notebook with microphone.

After installing Pyaudio and Speech_Recognition, there are still bugs. For example, missing PocketSphinx module: ensure that PocketSphinx is set up correctly can be solved by here

sudo pip install --upgrade pocketsphinx

I thought Anaconda will result some problems. If you can't solve your problems, try to de-active the anaconda.

uav-gesture-control_python's People

Contributors

Stargazers

Watchers

uav-gesture-control_python's Issues

video_effect lost response

I‘m so sorry that I have another problem. When I run the video_effect, it lost response and then stop runnning. So can you tell me what should I do? Thanks a lot.

關於使用openpose1.4.0依然出現問題

嗨！又是我在經過無數次測試之後我參考了Tame7790的問題，改用openpose1.4.0進行測試結果得到以下的結果

請問大大這個問題該如何解決？

備註：在cmake有句選Build_Python，OpenPose Demo也能順利運行

關於openpose與程式版本不同的問題

如題，我試過了很多辦法卻還是無法順利執行程式，我朋友指出問題為openpose版本不同所致，但作為一個程式新手實在不知道該如何修改程式。還請大老幫幫我。感謝

你好，请问你们这个应用，有发表相关论文可以拜读一下吗

AttributeError: 'module' object has no attribute 'OpenPose'

Hello, when I run python2 video_effect.py under the dir ~/openpose/build/examples/tutorial_api_python, there existed a problem
Traceback (most recent call last): File "video_effect.py", line 38, in <module> openpose = op.OpenPose(params) AttributeError: 'module' object has no attribute 'OpenPose'
the dir"tutorial_api_python" is the dir of your project"tutorial_python".
I have build openpose correctly and passed the test in quick start of https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/quick_start.md#quick-start .
Then I find an code named "openpose.py" in /home/yhh/openpose/build/examples/tutorial_api_python/python/openpose , where there defines an class object "OpenPose". How can I handle it to solve the AttributeError above?
Thank you and Thank you for your work.

一点问题

想请问如何执行video_effect.py这个档案，以及请将文件“ /python_tutorial”复制到计算机路径“ openpose / build / examples / python_tutorial”分别是指tutorial_python与tutorial_api_python吗?
又或者是我有遗漏掉了什么文件呢?

About Pose Detection by the Camera of Tello

Can you tell me the detailed steps, like after copy the files what should I do? My openpose can't use the tello's video. I can only open the openposeDemo and video.py Individually.

关于在windows环境下运行

请问下，可以在windows环境下运行吗，我安了python，但是点击了会闪退

How to apply the KNN to match gesture?

Thanks for your sharing.
Can you tell me the detail about the matching in gestures using KNN?I am a little confused
Thanks

Cmake error:The following variables are used in this project, but they are set to NOTFOUD

Hi, there,
When I use CMake GUI and make sure the souse code directory is /openpose, this error occured and all the mothods I gooled don't help.

My OS: win10

it's like this:
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
Caffe_LIB
linked by target "openpose" in directory C:/Tello/openpose/src/openpose
linked by target "Calibration" in directory C:/Tello/openpose/examples/calibration
linked by target "OpenPoseDemo" in directory C:/Tello/openpose/examples/openpose
linked by target "1_custom_post_processing" in directory C:/Tello/openpose/examples/tutorial_add_module
linked by target "01_body_from_image_default" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "02_whole_body_from_image_default" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "08_heatmaps_from_image" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "03_keypoints_from_image" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "12_asynchronous_custom_input_output_and_datum" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "05_keypoints_from_images_multi_gpu" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "04_keypoints_from_images" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "06_face_from_image" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "07_hand_from_image" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "09_keypoints_from_heatmaps" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "10_asynchronous_custom_input" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "11_asynchronous_custom_output" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "13_synchronous_custom_input" in directory C:/Tello/openpose/examples/tutorial_api_cpp
linked by target "17_synchronous_custom_all_and_datum" in directory

Thank you.

Feactures related to the selection of KNN

hi，First of all，thank you for sharing. I have some problems about choosing feature？
i found that your used （x2-x1),(y2-y1）/（|x2-x1| + |y2-y1|）as the feature. why this can distinguish between different pose ? I think this aboved parameters cound't represent.