Tensorflow目标检测接口配合tflite量化模型（二）

python export_inference_graph.py \
--input_type image_tensor \
--pipeline_config_path=ssd_mobilenet_v2_coco.config \   
--trained_checkpoint_prefix=training//model.ckpt-30000 \   
--output_directory models_trained/

import tensorflow as tf
import cv2
import os
import time
import numpy as np
from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util
 
videofile='/home/zyl/Documents/caffe/examples/MobileNet-SSD/videos/20180813140109903.avi'
cap=cv2.VideoCapture(videofile)
MODEL_NUM_CLASSES=3
MODEL_LABEL_MAP ='/home/zyl/data/dms_tf/label_map.pbtxt'
MODEL_PB='/home/zyl/data/dms_tf/model2/export_result/frozen_inference_graph.pb'
 
# read graph model
with tf.gfile.GFile(MODEL_PB,'rb') as fd:
    _graph=tf.GraphDef()
    _graph.ParseFromString(fd.read())
    tf.import_graph_def(_graph,name='')
    
# get the default graph model
detection_graph=tf.get_default_graph()
 
# read labelmap
label_map=label_map_util.load_labelmap(MODEL_LABEL_MAP)
categories=label_map_util.convert_label_map_to_categories(label_map,MODEL_NUM_CLASSES)
category_index=label_map_util.create_category_index(categories)
 
with tf.Session(graph=detection_graph) as sess:
    while(cap.isOpened()):
        ret,frame=cap.read()
        frame_np_expanded=np.expand_dims(frame,axis=0)
        image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
        boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
        scores = detection_graph.get_tensor_by_name('detection_scores:0')
        classes = detection_graph.get_tensor_by_name('detection_classes:0')
        num_detections = detection_graph.get_tensor_by_name('num_detections:0')
        t1=time.time()
        (boxes,scores,classes,num_detections)=sess.run([boxes,scores,classes,num_detections], \
        feed_dict={image_tensor:frame_np_expanded})
 
        vis_util.visualize_boxes_and_labels_on_image_array(frame,np.squeeze(boxes),
        np.squeeze(classes).astype(np.int32),np.squeeze(scores),category_index,
        use_normalized_coordinates=True,line_thickness=6)
        t2=time.time()
        print('FPS:',1/(t2-t1))
        cv2.imshow('MobilenetTF',frame)
        if cv2.waitKey(1)&0xff ==27:
            break
    cap.release()

sudo apt-get install pkg-config zip g++ zlib1g-dev unzip

chmod +x bazel-0.15.0-installer-linux-x86_64.sh
./bazel-0.15.0-installer-linux-x86_64.sh --user

vim ~/.bashrc

export PATH="$PATH:$HOME/bin"

mkdir C:\tensorflow-build
cd C:\tensorflow-build
git clone https://github.com/tensorflow/tensorflow.git 
cd tensorflow 
git checkout r1.13   # 签出TensorFlow v1.13的分支

python ./configure.py

yhilly@ubuntu:~/tensorflow$ ./configure 
You have bazel 0.9.0 installed.
Please specify the location of python. [Default is /home/yhilly/anaconda3/envs/tensorflow/bin/python]:

Do you wish to build TensorFlow with jemalloc as malloc support? [Y/n]: 
jemalloc as malloc support will be enabled for TensorFlow.

Do you wish to build TensorFlow with Google Cloud Platform support? [Y/n]: n
No Google Cloud Platform support will be enabled for TensorFlow.

Do you wish to build TensorFlow with Hadoop File System support? [Y/n]: n
No Hadoop File System support will be enabled for TensorFlow.

Do you wish to build TensorFlow with XLA JIT support? [y/N]: n
No XLA JIT support will be enabled for TensorFlow.

Do you wish to build TensorFlow with CUDA support? [y/N]: 
No CUDA support will be enabled for TensorFlow.

bazel build --config=opt //tensorflow/tools/pip_package:build_pip_package

bazel build --config=opt --config=cuda --define=no_tensorflow_py_deps=true //tensorflow/tools/pip_package:build_pip_package

bazel-bin/tensorflow/tools/pip_package/build_pip_package  ~/tensorflow_pkg

pip  install  ~/tensorflow_pkg/<Paste full .whl filename here>.whl

python

>>> import tensorflow as tf
>>> tf.__version__

python export_tflite_ssd_graph.py \
    --pipeline_config_path=ssd_mobilenet_v2_coco.config \
    --trained_checkpoint_prefix=models_trained/model.ckpt-30000\
    --output_directory=models

bazel-bin/tensorflow/lite/toco/toco \
--input_file=/home/zyl/Documents/tfmodels/dms/models/tflite_graph.pb \
--input_format=TENSORFLOW_GRAPHDEF  \
--output_format=TFLITE  \
--output_file=/home/zyl/Documents/tfmodels/dms/models/litefloat_zyl.tflite \
--inference_type=FLOAT \
--input_arrays=normalized_input_image_tensor \
--output_arrays='TFLite_Detection_PostProcess','TFLite_Detection_PostProcess:1','TFLite_Detection_PostProcess:2','TFLite_Detection_PostProcess:3' \
--input_shapes=1,300,300,3 \
--mean_values=128 \
--std_dev_values=128 \
--default_ranges_min=0 \
--allow_custom_ops

bazel-bin/tensorflow/lite/toco/toco \
--graph_def_file=/home/zyl/data/dms_tf/model1/tflite_graph.pb  \
--output_file=/home/zyl/data/dms_tf/model1/tflite_model/model1.tflite \
--input_shapes=1,300,300,3 \
--input_arrays=normalized_input_image_tensor \
--output_arrays='TFLite_Detection_PostProcess','TFLite_Detection_PostProcess:1','TFLite_Detection_PostProcess:2','TFLite_Detection_PostProcess:3' \
--inference_type=QUANTIZED_UINT8 \
--mean_values=128 \
--std_dev_values=128 \
--default_ranges_min=0 \
--default_ranges_max=6 \
--change_concat_input_ranges=False \
--allow_custom_ops

wget https://raw.githubusercontent.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/master/TFLite_detection_image.py --no-check-certificate
wget https://raw.githubusercontent.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/master/TFLite_detection_video.py --no-check-certificate
wget https://raw.githubusercontent.com/EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/master/TFLite_detection_webcam.py --no-check-certificate

python TFLite_detection_image.py -h