YOLO+yolo9000配置使用darknet

• Installing Darknet

1.直接设置使用，编译通过

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

2. 下载权重测试

    wget http://pjreddie.com/media/files/yolo.weights
    ./darknet yolo test cfg/yolo.cfg yolo.weights data/dog.jpg  

3.测试结果：

dsp@dsp:/media/dsp/学习/yolo_darknet/darknet$ ./darknet yolo test cfg/yolo.cfg yolo.weights data/dog.jpg
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 9.566333 seconds.
Not compiled with OpenCV, saving to predictions.png instead

- 区分上下连个命令：

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

-/darknet detect cfg/yolo.cfg yolo.weights data/dog.jpg 结果：

4.设置GPU和opencv

GPU=
OPENCV=

-编译错误

/usr/bin/ld: cannot find -lippicv
collect2: error: ld returned  exit status
Makefile:: recipe for target 'darknet' failed
make: *** [darknet] Error
/usr/bin/ld: cannot find -make: *** Waiting for unfinished jobs....
lippicv
collect2: error: ld returned  exit status
Makefile:: recipe for target 'libdarknet.so' failed
make: *** [libdarknet.so] Error 

- opencv安装的问题，于是重新安装

https://github.com/pjreddie/darknet/issues/290:  /usr/bin/ld: cannot find -lippicv collect2: error: ld returned 1 exit status

https://github.com/opencv/opencv/issues/5852这篇文章中也遇到了这个错误. 可以在cmake时加上"cmake -DINSTALL_CREATE_DISTRIB=ON" 这句,就不会有这个错误了.

- 首先卸载opencv，然后在安装 ：卸载参考：Ubuntu16.04 上openCV的卸载与 opencv3.0.2安装记录

cd /home/hy/opencv/build
make uninstall//卸载掉配置路径中的文件
sudo rm -r build//删除build文件
//删除掉环境中有关的其余包
sudo rm -r /usr/local/include/opencv2 /usr/local/include/opencv /usr/include/opencv /usr/include/opencv2 /usr/local/share/opencv /usr/local/share/OpenCV /usr/share/opencv /usr/share/OpenCV /usr/local/bin/opencv* /usr/local/lib/libopencv*

-安装，安装有图割错误，以前解决办法

dsp@dsp:~/opencv-3.1./build$ sudo cmake -D CMAKE_BUILD_TYPE=Release -D CMAKE_INSTALL_PREFIX=/usr/local .. -DWITH_IPP=ON -DINSTALL_CREATE_DISTRIB=ON

dsp@dsp:~/opencv-3.1./build$ sudo make -j$(nproc)

5. 在编译darknet

dart -lcublas -lcurand -lcudnn -lstdc++  libdarknet.a
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFReadRGBAStrip@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFReadDirectory@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFWriteEncodedStrip@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFIsTiled@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFWriteScanline@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFGetField@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFNumberOfStrips@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFScanlineSize@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFReadEncodedTile@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFReadRGBATile@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFClose@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFRGBAImageOK@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFOpen@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFReadEncodedStrip@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFSetField@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFSetWarningHandler@LIBTIFF_4.'
/usr/local/lib/libopencv_imgcodecs.so: undefined reference to `TIFFSetErrorHandler@LIBTIFF_4.'
collect2: error: ld returned  exit status
Makefile:: recipe for target 'darknet' failed
make: *** [darknet] Error 

//权限问题，sudo即可解决
这个可能是权限问题，采用以下指令：
sudo su；
一切都能顺利解决。

6. 运行yolo和yolo9000，可以实现demo运行，并实现视频检测

• philipperemy/yolo-9000

YOLOv2是Joseph Redmon提出的针对YOLO算法不足的改进版本，作者使用了一系列的方法对原来的YOLO多目标检测框架进行了改进，在保持原有速度的优势之下，精度上得以提升，此外作者提出了一种目标分类与检测的联合训练方法，通过这种方法YOLO9000可以同时在COCO和ImageNet数据集中进行训练，训练后的模型可以实现多达9000种物体的实时检测。

• 第一种：为darknet添加Python接口

  Github：https://github.com/SidHard/py-yolo2

• 第二种：使用keras

  Github：https://github.com/allanzelener/YAD2K