利用caffe的pycaffe接口将caffemodel中的字段保存为prototxt

使用str可以一次性将caffemodel中所有的的字段读出来
例如：

# coding:utf-8

import _init_paths
import caffe.proto.caffe_pb2 as caffe_pb2

caffemodel_filename = 'F:/caffe_new/caffe/caffe-master/data/mnist-test/mnist-models/lenet_iter_10000.caffemodel'

model = caffe_pb2.NetParameter()

f = open(caffemodel_filename, 'rb')
model.ParseFromString(f.read())
f.close()

print model.__str__

该代码已经可以打印出网络训练所需的train.prototxt，但同时也会将blobs中的不必要的data数据段也打印出来，这个字段里面有太多的内容，是经过多次迭代学习出来的卷积核以及bias的数值。这些字段在prototxt文件中应当忽略。还有str输出的首尾也有不必要的字符串也要去掉。

我们可以将str读出的字段输出到文件，并将不必要的字段去掉。
因为解析caffemode时会读到大量的blobs占用很长时间，没有使用大的caffemodel，所以该代码是以lenet_iter_10000.caffemodel为例
下面是处理后的代码

# coding:utf-8

#import _init_paths
import caffe.proto.caffe_pb2 as caffe_pb2

caffemodel_filename = 'F:/caffe_new/caffe/caffe-master/data/mnist-test/mnist-models/lenet_iter_10000.caffemodel'
#caffemodel_filename = 'F:/py-faster-rcnn-master/data/imagenet_models/ZF.v2.caffemodel'

model = caffe_pb2.NetParameter()

f = open(caffemodel_filename, 'rb')
model.ParseFromString(f.read())
f.close()

import sys
old=sys.stdout
save_filename = 'lenet_from_caffemodel.prototxt'
sys.stdout=open( save_filename, 'w')
print model.__str__
sys.stdout=old
f.close()

#linux下的sed命令可以对文本中的字段进行操作，而我们使用的为Windows系统
#import os
#cmd_1 = 'sed -i "1s/^.\{38\}//" ' + save_filename     # 删除第一行前面38个字符
#cmd_1 = 'sed -i "1s/^" ' + save_filename     # 删除第一行前面38个字符
#cmd_2 = "sed -i '$d' " + save_filename      # 删除最后一行

#os.system(cmd_1)
#os.system(cmd_2)

# 打开刚刚存储的文件，输出里面的内容，输出时过滤掉“blobs”块和"phase: TRAIN"行。
f=open(save_filename, 'r')
lines = f.readlines() #readlines()自动将文件内容分析成一个行的列表，该列表可以由 python 的 for... in ... 结构进行处理
f.close()
wr = open(save_filename, 'w')
now_have_blobs = False
nu = 1
for line in lines:
    #print nu
    nu = nu + 1
    content = line.strip('\n') #strip(rm)函数的作用为删除line字符串中开头、结尾处，位于rm删除序列的字符(即所有'\n')

    if (content == '  blobs {'): #如果字段包含'{'则开始
        now_have_blobs = True
    elif (content == '  }' and now_have_blobs==True): #结束的条件
        now_have_blobs = False
        continue

    if (content == '  phase: TRAIN'): #不保留bolbs中的  phase: TRAIN字段
        continue

    if (now_have_blobs):
        continue
    else:
        wr.write(content+'\n')  #满足条件则将字段写入
wr.close()

下面为生成的lenet_from_caffemodel.prototxt

<bound method NetParameter.__str__ of name: "LeNet" 
layer {
  name: "mnist"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
    scale: 0.00390625
    mean_file: "mnist-data/trainMean.binaryproto"
  }
  data_param {
    source: "mnist-data/mtrainlmdb"
    batch_size: 50
    backend: LMDB
  }
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1.0
  }
  param {
    lr_mult: 2.0
  }
  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: 1.0
  }
  param {
    lr_mult: 2.0
  }
  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: 1.0
  }
  param {
    lr_mult: 2.0
  }
  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: 1.0
  }
  param {
    lr_mult: 2.0
  }
  inner_product_param {
    num_output: 10
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "loss"
  type: "SoftmaxWithLoss"
  bottom: "ip2"
  bottom: "label"
  top: "loss"
  loss_weight: 1.0
}
>

由于还未找到删除文本首行和尾行的windows命令，暂时还不知道怎么处理这两行的字段
如果手动去除首行和尾行不必要的字符，再利用netscope工具画出的网络结构图为

参考：
http://www.cnblogs.com/zjutzz/p/6185452.html