TJArk Code Release 2019

Introduction

This is the open source code of TJArk framework used on NAO V6 robots in RoboCup 2019.

Architecture

• Bin: executable binaries of the code
• Config: configuration files
• Src: source code

Features

• Five threads are used: Upper-cognition, Lower-cognition, Motion, Behavior, Debug
• Optimized version of OpenCV

Some third-party libraries/codes are included in this code release, such as LolaConnector from Nao-Team HTWK, CABSL(iostream version) from B-Human, and OpenCV.

Their license can be found in 3rdPartyLicense folder. We highly respect and appreciate their contribution.

How to use on NAO v6

• Use installCrossToolChain.sh to install cross toolchain

• Run path-to-toolchain/yocto-sdk/relocate_qitoolchain.sh

  Caution: After this the toolchain can't be moved or copied to other devices anymore!

• Run create_crossconfig.py to create cross-config.cmake. This is the cmake cross chain file.

• Run ctc/yocto-sdk/relocate_sdk.py to relocate sdk.

• Run sh Make.sh to build this program.

If you are interested in other modules or our dataset or pre-trained weights, please contact us TJArk.Official@gmail.com.