配置YOLO2（ubuntu16.04+cuda8.0+opencv3.1.0）

要求已经安装好了CUDA 8.0 以及OpenCV3.1.0

YOLO官方网站

配置darknet

git clone https://github.com/pjreddie/darknet
cd darknet
make  

如果没有报错输入

./darknet  

得到输出

./darknet <function>  

说明darknet配置成功

打开Makefile文件，将开头几行改为

GPU=
CUDNN=
OPENCV=

接着查询自己GPU的计算能力，查看Makefile是否包含，如下我的GPU计算能力为6.1,调整后为：

ARCH= -gencode arch=compute_30,code=sm_30 \
      -gencode arch=compute_35,code=sm_35 \
      -gencode arch=compute_50,code=[sm_50,compute_50] \
      -gencode arch=compute_61,code=[sm_61,compute_61] \
      -gencode arch=compute_52,code=[sm_52,compute_52]

之后重新编译，就可以实现基于cuda和opencv的编译

下载预训练文件

wget https://pjreddie.com/media/files/yolo.weights  

测试

./darknet detect cfg/yolo.cfg yolo.weights data/dog.jpg  

我得到的结果是：

layer     filters    size              input                output
     conv        x  /     x  x      ->    x  x
     max           x  /     x  x     ->    x  x
     conv        x  /     x  x     ->    x  x
     max           x  /     x  x     ->    x  x
     conv       x  /     x  x     ->    x  x
     conv        x  /     x  x    ->    x  x
     conv       x  /     x  x     ->    x  x
     max           x  /     x  x    ->     x   x
     conv       x  /      x   x    ->     x   x
     conv       x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    max           x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    conv       x  /      x   x    ->     x   x
    max           x  /      x   x    ->     x   x
    conv      x  /      x   x    ->     x   x1024
    conv       x  /      x   x1024   ->     x   x
    conv      x  /      x   x    ->     x   x1024
    conv       x  /      x   x1024   ->     x   x
    conv      x  /      x   x    ->     x   x1024
    conv      x  /      x   x1024   ->     x   x1024
    conv      x  /      x   x1024   ->     x   x1024
    route
    conv        x  /      x   x    ->     x   x
    reorg              /      x   x     ->     x   x
    route
    conv      x  /      x   x1280   ->     x   x1024
    conv       x  /      x   x1024   ->     x   x
    detection
mask_scale: Using default '1.000000'
Loading weights from yolo.weights...Done!
data/dog.jpg: Predicted in 0.070790 seconds.
dog: %
car: %
truck: %
bicycle: %
init done
opengl support available