baidu / boteye_sensor

boteye_sensor

This repo contains hardware and firmware of Baidu boteye sensor. If you purchase the hardware from Kinco Automation(a third party manufacturer), which is already loaded with the latest firmware, you can skip the following steps.

Hardware Introduction

Firmware supports three hardware versions: XP, XP2, and XP3.

FX3 SDK Installation

In order to compile, you need cypress FX3 sdk (http://www.cypress.com/documentation/software-and-drivers/ez-usb-fx3-software-development-kit).

Download the sdk here http://www.cypress.com/file/221546

You can get fx3_firmware_linux.tar.gz and cyusb_linux_1.0.4.tar.gz, then extract to root directory of this repo.

• Install arm gcc compiler

  sudo apt-get install gcc-arm-none-eabi

• Make sure your arm compiler is arm-none-eabi-gcc version 4.8.2

  arm-none-eabi-gcc --version

• Install gcc compiler

  sudo apt-get install build-essential

• Modify SDK config file

  • Configure build type in the file cyfx3sdk/fw_build/fx3_fw/fx3_build_config.mak

    Change CYCONFOPT=fx3_debug to CYCONFOPT=fx3_release

  • Configure the arm compiler info in the file cyfx3sdk/fw_build/fx3_fw/fx3_armgcc_config.mak

    Change LDLIBS += section to your lib locations, most likely located in /usr/arm-linux-gnueabi/lib/libc.a and /usr/lib/gcc/arm-none-eabi/4.8.2/libgcc.a, and replace the default like this:

    The ARM toolchain location and the version are taken from environment variables
    LDLIBS  += \
    /usr/lib/arm-none-eabi/lib/libc.a \
    /usr/lib/gcc/arm-none-eabi/4.8.2/libgcc.a \
    
    EXEEXT		= elf

• Compile img format convert tools: elf2img

  To convert elf to img file, you need to compile cyfx3sdk/util/elf2img/elf2img.c

  cd cyfx3sdk/util/elf2img
  gcc elf2img.c -o elf2img -O3

Compile Firmware

• If everything is all set, you can run ./build.sh within boteye_sensor/firmware, which generates cyfxuvc.img.

FX3 Image Download Tool Install

• Install libusb-dev

  sudo apt-get install libusb-1.0-0-dev

• Install FX3 tools, please read the README in ./cyusb_linux_1.0.4.

  cd cyusb_linux_1.0.4
  make
  chmod +x install.sh
  sudo ./install.sh

• If you find an error qmake-qt4: command not found, please install qt4 by running

  sudo apt-get install qt4-dev-tools qt4-doc qt4-qtconfig qt4-demos qt4-designer

• Add execute premission of cyusb_linux

  sudo chmod +x /usr/local/bin/cyusb_linux

• Add the following two lines to ~/.bashrc, run source ~/.bashrc, and assign CYUSB_ROOT to the corresponding file path.

  export CYUSB_ROOT=~/Development/sensor/cyusb_linux_1.0.4/
  export LD_LIBRARY_PATH=/usr/local/lib/
  source ~/.bashrc

Image Download

• Use cyusb_linux to flash your img file to the board flash. If the sensor has an img on flash, it will go into auto-run mode with red LED blinking for 1s. If the sensor doesn't have an img on flash, the red LED will be always on. We can use ./sensor/host_bin/spi_test to erase the img on flash so that we can download a new one.
• example: ./sensor/host_bin/spi_test /dev/video0

Firmware Tools

• There are two tools you can use in firmware/host_bin.
  • spi_test: erase flash
  • version_test: read hardware and software version.
• You can also install luvcview or other camera view tools to test the camera. However, you will most likely see green and red ghost images due to the custom formatting of the stereo raw images.

About Version

• Hardware version:
  • You can use command dmesg to check USB Product attribute, it will show the hardware version.
  • Example: Baidu_Robotics_vision_XP2
• Software version:
  • You can also use dmesg to check USB SeriaNumber attribute, it will show the software version.
  • Example: V0.3.7-839610e-commit [version num - commit hash - commit/dirty]