tensorflow object detection模块

本文只是介绍一下tensorflow object detection模块的安装与使用。
1、下载tensorflow 的models模块https://github.com/tensorflow/models.git

2、编译Protobuf库，在object_detection同级目录下打开终端运行：

      protoc object_detection/protos/*.proto --python_out=.

3、在.bashrc加入环境变量

      export PYTHONPATH=$PYTHONPATH:"/home/han/models/research:/home/han/models/research/slim"

然后source ~/.bashrc

4、在object_detection同级目录下检测是否成功
python object_detection/builders/model_builder_test.py

5、数据预处理，将数据转为TFRecord格式，这里测试采用的VOC2012数据集。将数据集解压到/home/han/DataSets中
采用如下命令：
训练集

python object_detection/dataset_tools/create_pascal_tf_record.py --label_map_path=object_detection/data/pascal_label_map.pbtxt --data_dir=/home/han/DataSets/VOCdevkit-2012 --year=VOC2012 --set=train --output_path=/home/han/DataSets/pascal_train.record

验证集

python object_detection/dataset_tools/create_pascal_tf_record.py --label_map_path=object_detection/data/pascal_label_map.pbtxt  --data_dir=/home/han/DataSets/VOCdevkit-2012 --year=VOC2012 --set=val --output_path=/home/han/DataSets/pascal_val.record

6、在object_detection文件夹下新建文件夹train2012，将生成的pascal_train.record和pacal_val.record放在此文件夹下，同时将data/pascal_label_map.pbtxt文件拷贝到train2012

7、新建trainmodels文件夹，解压下载的ssd_mobilenet_v1_coco_11_06_2017.tar.gz，将里面的model.ckpt*的三个文件拷贝到trainmodels。

将文件object_detection/samples/configs/ssd_mobilenet_v1_pets.config复制到trainmodels，打开做如下修改：

（1）num_classes:修改为自己的classes num ，这里为20

（2）将所有PATH_TO_BE_CONFIGURED的地方类比修改为自己之前设置的路径（5处），按照自己的路径修改即可

# SSD with Mobilenet v1, configured for Oxford-IIIT Pets Dataset.
# Users should configure the fine_tune_checkpoint field in the train config as
# well as the label_map_path and input_path fields in the train_input_reader and
# eval_input_reader. Search for "PATH_TO_BE_CONFIGURED" to find the fields that
# should be configured.

model {
  ssd {
    num_classes: 20
    box_coder {
      faster_rcnn_box_coder {
        y_scale: 10.0
        x_scale: 10.0
        height_scale: 5.0
        width_scale: 5.0
      }
    }
    matcher {
      argmax_matcher {
        matched_threshold: 0.5
        unmatched_threshold: 0.5
        ignore_thresholds: false
        negatives_lower_than_unmatched: true
        force_match_for_each_row: true
      }
    }
    similarity_calculator {
      iou_similarity {
      }
    }
    anchor_generator {
      ssd_anchor_generator {
        num_layers: 6
        min_scale: 0.2
        max_scale: 0.95
        aspect_ratios: 1.0
        aspect_ratios: 2.0
        aspect_ratios: 0.5
        aspect_ratios: 3.0
        aspect_ratios: 0.3333
      }
    }
    image_resizer {
      fixed_shape_resizer {
        height: 300
        width: 300
      }
    }
    box_predictor {
      convolutional_box_predictor {
        min_depth: 0
        max_depth: 0
        num_layers_before_predictor: 0
        use_dropout: false
        dropout_keep_probability: 0.8
        kernel_size: 1
        box_code_size: 4
        apply_sigmoid_to_scores: false
        conv_hyperparams {
          activation: RELU_6,
          regularizer {
            l2_regularizer {
              weight: 0.00004
            }
          }
          initializer {
            truncated_normal_initializer {
              stddev: 0.03
              mean: 0.0
            }
          }
          batch_norm {
            train: true,
            scale: true,
            center: true,
            decay: 0.9997,
            epsilon: 0.001,
          }
        }
      }
    }
    feature_extractor {
      type: 'ssd_mobilenet_v1'
      min_depth: 16
      depth_multiplier: 1.0
      conv_hyperparams {
        activation: RELU_6,
        regularizer {
          l2_regularizer {
            weight: 0.00004
          }
        }
        initializer {
          truncated_normal_initializer {
            stddev: 0.03
            mean: 0.0
          }
        }
        batch_norm {
          train: true,
          scale: true,
          center: true,
          decay: 0.9997,
          epsilon: 0.001,
        }
      }
    }
    loss {
      classification_loss {
        weighted_sigmoid {
          anchorwise_output: true
        }
      }
      localization_loss {
        weighted_smooth_l1 {
          anchorwise_output: true
        }
      }
      hard_example_miner {
        num_hard_examples: 3000
        iou_threshold: 0.99
        loss_type: CLASSIFICATION
        max_negatives_per_positive: 3
        min_negatives_per_image: 0
      }
      classification_weight: 1.0
      localization_weight: 1.0
    }
    normalize_loss_by_num_matches: true
    post_processing {
      batch_non_max_suppression {
        score_threshold: 1e-8
        iou_threshold: 0.6
        max_detections_per_class: 100
        max_total_detections: 100
      }
      score_converter: SIGMOID
    }
  }
}

train_config: {
  batch_size: 24
  optimizer {
    rms_prop_optimizer: {
      learning_rate: {
        exponential_decay_learning_rate {
          initial_learning_rate: 0.004
          decay_steps: 800720
          decay_factor: 0.95
        }
      }
      momentum_optimizer_value: 0.9
      decay: 0.9
      epsilon: 1.0
    }
  }
  fine_tune_checkpoint: "/home/han/models/research/object_detection/trainmodels/model.ckpt"
  from_detection_checkpoint: true
  # Note: The below line limits the training process to 200K steps, which we
  # empirically found to be sufficient enough to train the pets dataset. This
  # effectively bypasses the learning rate schedule (the learning rate will
  # never decay). Remove the below line to train indefinitely.
  num_steps: 200000
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
  data_augmentation_options {
    ssd_random_crop {
    }
  }
}

train_input_reader: {
  tf_record_input_reader {
    input_path: "/home/han/models/research/object_detection/train2012/pascal_train.record"
  }
  label_map_path: "/home/han/models/research/object_detection/train2012/pascal_label_map.pbtxt"
}

eval_config: {
  num_examples: 2000
  # Note: The below line limits the evaluation process to 10 evaluations.
  # Remove the below line to evaluate indefinitely.
  max_evals: 10
}

eval_input_reader: {
  tf_record_input_reader {
    input_path: "/home/han/models/research/object_detection/train2012/pascal_val.record"
  }
  label_map_path: "/home/han/models/research/object_detection/train2012/pascal_label_map.pbtxt"
  shuffle: false
  num_readers: 1
}

8、在object_detection目录下执行如下命令

python train.py --train_dir=train2012 --pipeline_config_path=trainmodels/ssd_mobilenet_v1_pets.config

9、生成pb文件

新建一个pb文件夹，将train2012文件夹下选择如下文件放到pb文件夹内

checkpoint

model.ckpt.data-00000-of-00001

model.ckpt.index

model.ckpt.meta

去掉ckpt后面的数字

建立一个sh 脚本

#!/bin/bash

CONFIG_PATH=./trainmodels/ssd_mobilenet_v1_pets.config
MODEL_PATH=./pb/model.ckpt
SAVE_DIR=./pb

python export_inference_graph.py \
--pipeline_config_path ${CONFIG_PATH} \
--trained_checkpoint_prefix ${MODEL_PATH} \
--output_directory ${SAVE_DIR}