caffe 从mnsit示例到自己创建数据集再到识别手写图片

记录下使用caffe mnist数据集，训练+识别图片过程中遇到的问题。
很不详细。

1. 按部就班，获取mnist数据集、创建lmdb、使用example文件训练，得到caffemodel，找了图片预测，错误率很高
   原因： mnist原始图片是黑底白字，找的图片是白底黑字。。。自己来重做lmdb吧。—注意灰度图的格式
   2.网络上找来mnist手写图片，白底黑字，使用脚本生成lmdb

cp create_imagenet.sh create_imagenet_my_mnist.sh
cat create_imagenet_my_mnist.sh
#!/usr/bin/env sh
# Create the imagenet lmdb inputs
# N.B. set the path to the imagenet train + val data dirs

EXAMPLE=examples/mnist
DATA=data/mnist/mnist
TOOLS=build/tools

TRAIN_DATA_ROOT=/root/caffe-0.16.3/data/mnist/mnist/train/
VAL_DATA_ROOT=/root/caffe-0.16.3/data/mnist/mnist/test/

# Set RESIZE=true to resize the images to 256x256. Leave as false if images have
# already been resized using another tool.
RESIZE=true
if $RESIZE; then
  RESIZE_HEIGHT=28
  RESIZE_WIDTH=28
else
  RESIZE_HEIGHT=0
  RESIZE_WIDTH=0
fi

if [ ! -d "$TRAIN_DATA_ROOT" ]; then
  echo "Error: TRAIN_DATA_ROOT is not a path to a directory: $TRAIN_DATA_ROOT"
  echo "Set the TRAIN_DATA_ROOT variable in create_imagenet.sh to the path" \
       "where the ImageNet training data is stored."
  exit 1
fi

if [ ! -d "$VAL_DATA_ROOT" ]; then
  echo "Error: VAL_DATA_ROOT is not a path to a directory: $VAL_DATA_ROOT"
  echo "Set the VAL_DATA_ROOT variable in create_imagenet.sh to the path" \
       "where the ImageNet validation data is stored."
  exit 1
fi

echo "Creating train lmdb..."

GLOG_logtostderr=1 $TOOLS/convert_imageset \
    --resize_height=$RESIZE_HEIGHT \
    --resize_width=$RESIZE_WIDTH \
    --shuffle \
    --gray \
    $TRAIN_DATA_ROOT \ 
    $DATA/train.txt \  ##注意如果图片就在$DATA下面，train.txt和val.txt里的路径$DATA下的路径就可以
    $EXAMPLE/my_mnist_train_lmdb

echo "Creating val lmdb..."

GLOG_logtostderr=1 $TOOLS/convert_imageset \
    --resize_height=$RESIZE_HEIGHT \
    --resize_width=$RESIZE_WIDTH \
    --shuffle \
    --gray \  ###默认是RGB，需要加个强制灰度，否则预测依然不准
    $VAL_DATA_ROOT \
    $DATA/val.txt \
    $EXAMPLE/my_mnist_val_lmdb

echo "Done."
###开始识别
name: "LeNet"
layer {
  name: "data"
  type: "Input"
  top: "data"
  input_param { shape: { dim: 1 dim: 1 dim: 28 dim: 28 } }
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 20
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "conv1"
  top: "pool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv2"
  type: "Convolution"
  bottom: "pool1"
  top: "conv2"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 50
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "pool2"
  type: "Pooling"
  bottom: "conv2"
  top: "pool2"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "ip1"
  type: "InnerProduct"
  bottom: "pool2"
  top: "ip1"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  inner_product_param {
    num_output: 500
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "relu1"
  type: "ReLU"
  bottom: "ip1"
  top: "ip1"
}
layer {
  name: "ip2"
  type: "InnerProduct"
  bottom: "ip1"
  top: "ip2"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  inner_product_param {
    num_output: 10
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "prob"
  type: "Softmax"
  bottom: "ip2"
  top: "prob"
}


[root@7ed2726e0c5a caffe-0.16.3]# ./build/examples/cpp_classification/classification.bin examples/mnist/lenet_deploy.prototxt examples/mnist/lenet_iter_10000.caffemodel examples/mnist/mean.binaryproto data/mnist/mnist/test/labels.txt /root/8.png
---------- Prediction for /root/8.png ----------
1.0000 - "8"
0.0000 - "5"
0.0000 - "3"
0.0000 - "1"
0.0000 - "4"
##导出镜像，有空再继续

遇到的几个小坑：
1. 可能是第一次弄不太熟悉，需要保证训练集-模型-测试集，图片格式的一致
2. 注意均值文件，训练用了，预测的时候不能忘用
3. 创建lmdb时 灰度图不是默认选项，需要加参数