autoware入门教程-安装ENet

autoware入门教程-安装ENet

说明：

• 介绍如何在autoware下车安装ENet
• 测试环境：jetpack3.3

步骤：

• 安装相关依赖，参考说明

• 下载源码

cd ~
git clone --recursive https://github.com/TimoSaemann/ENet.git
cd ENet/caffe-enet

• 配置Makeconfig

cp Makefile.config.example Makefile.config

• 配置内容：

USE_OPENCV := 1
OPENCV_VERSION := 3
CUDA_ARCH := -gencode arch=compute_30,code=sm_30 \
             -gencode arch=compute_35,code=sm_35 \
             -gencode arch=compute_50,code=sm_50 \
             -gencode arch=compute_52,code=sm_52 \
             -gencode arch=compute_61,code=sm_61

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial/

• 编译

make && make distribute

• 下载weights文件，参考地址

• 解压cityscapes_weights.caffemodel文件到目录/home/ubuntu/ENet/enet_weights_zoo

• 运行

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:~/ENet/caffe-enet/distribute/lib
roslaunch vision_segment_enet_detect vision_segment_enet_detect.launch

• 如果没有image_segmenter_enet.launch文件，就需要重新编译一次autoware，参考
• 删除build和install里面的image_segmenter_enet,单独编译vision_segment_enet_detect

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:~/ENet/caffe-enet/distribute/lib
cd ~/autoware.ai 
AUTOWARE_COMPILE_WITH_CUDA=1 colcon build --cmake-args -DCMAKE_BUILD_TYPE=Release --packages-select vision_segment_enet_detect

• 如果没有出错，即说明配置完成

参考：

• https://gitlab.com/autowarefoundation/autoware.ai/core_perception/tree/master/vision_segment_enet_detect