SSD中ssd_detect调用caffe解读

SSD目录下，examples/ssd/下有一ssd_detect.cpp文件，编译连接后生成ssd_deetct.bin可以加以调用，该文件主要是对cpp_classification.cpp的改写，是一个利用C++调用Caffe的主体步骤，下面对其加以解读。

// This is a demo code for using a SSD model to do detection.
// The code is modified from examples/cpp_classification/classification.cpp.
// Usage:
//    ssd_detect [FLAGS] model_file weights_file list_file
//
// where model_file is the .prototxt file defining the network architecture, and
// weights_file is the .caffemodel file containing the network parameters, and
// list_file contains a list of image files with the format as follows:
//    folder/img1.JPEG
//    folder/img2.JPEG
// list_file can also contain a list of video files with the format as follows:
//    folder/video1.mp4
//    folder/video2.mp4
//
#include <caffe/caffe.hpp>
#ifdef USE_OPENCV
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#endif  // USE_OPENCV
#include <algorithm>
#include <iomanip>
#include <iosfwd>
#include <memory>
#include <string>
#include <utility>
#include <vector>

#ifdef USE_OPENCV
using namespace caffe;  // NOLINT(build/namespaces)
//定义检测器
class Detector {
public:
Detector(const string& model_file,
const string& weights_file,
const string& mean_file,
const string& mean_value);

std::vector<vector<float> > Detect(const cv::Mat& img);

private:
void SetMean(const string& mean_file, const string& mean_value);  //对mean_进行初始化

void WrapInputLayer(std::vector<cv::Mat>* input_channels);        //将 input_channels与网络输入绑定

void Preprocess(const cv::Mat& img,
std::vector<cv::Mat>* input_channels);           //一系列的预处理

private:
shared_ptr<Net<float> > net_;                                   //网络指针
cv::Size input_geometry_;                                       //输入图片的size
int num_channels_;                                             //输入图片的channel
cv::Mat mean_;                                                 //均值图片
};
//构造函数
Detector::Detector(const string& model_file,
const string& weights_file,
const string& mean_file,
const string& mean_value) {
#ifdef CPU_ONLY                         //定义工作模式CPU or GPU
Caffe::set_mode(Caffe::CPU);
#else
Caffe::set_mode(Caffe::GPU);
#endif

/* Load the network. */
net_.reset(new Net<float>(model_file, TEST));                        //从model_file中读取网络结构，初始化网络
net_->CopyTrainedLayersFrom(weights_file);                                     //从权值文件中读取网络参数，初始化net_

CHECK_EQ(net_->num_inputs(), 1) << "Network should have exactly one input.";
CHECK_EQ(net_->num_outputs(), 1) << "Network should have exactly one output.";

Blob<float>* input_layer = net_->input_blobs()[0];
num_channels_ = input_layer->channels();                                      //读取输入图片的channel
CHECK(num_channels_ == 3 || num_channels_ == 1)
<< "Input layer should have 1 or 3 channels.";
input_geometry_ = cv::Size(input_layer->width(), input_layer->height());      //读取输入图片的size

/* Load the binaryproto mean file. */
SetMean(mean_file, mean_value);                                               //初始化均值图片，至此所有成员变量都被初始化了
}

std::vector<vector<float> > Detector::Detect(const cv::Mat& img) {                       //检测器，输入图片，返回结果，每个vector代表一个结果，

//存有位置及信任程度信息
Blob<float>* input_layer = net_->input_blobs()[0];
input_layer->Reshape(1, num_channels_,
input_geometry_.height, input_geometry_.width);
/* Forward dimension change to all layers. */
net_->Reshape();                                                            //对网络进行reshape

std::vector<cv::Mat> input_channels;
WrapInputLayer(&input_channels);                                         //这地方比较有意思，将网络输入与input_channels绑定 ，后面再提
Preprocess(img, &input_channels);                                       //一些列预处理，后面详解

net_->Forward();                                                       //网络的前向传播

/* Copy the output layer to a std::vector */
Blob<float>* result_blob = net_->output_blobs()[0];
const float* result = result_blob->cpu_data();                          //读取输出信息
const int num_det = result_blob->height();                              //由此可见，只用了最后最后两维存储信息（只有一张图片，到现在没搞清前两维存什么）

//height：检测到的数量。width：检测到的每个目标的信息
vector<vector<float> > detections;
for (int k = 0; k < num_det; ++k) {
if (result[0] == -1) {                                            //-1代表是背景
// Skip invalid detection.
result += 7;                                                          //由此可见。每个目标占7个位置，后面解释每个位置的意思
continue;
}
vector<float> detection(result, result + 7);                            //vector的构造方法之一
detections.push_back(detection);
result += 7;
}
return detections;
}

/* Load the mean file in binaryproto format. */
void Detector::SetMean(const string& mean_file, const string& mean_value) {                //设置均值文件，Caffe中有两种设置均值文件的方式，

//mean_file or mean_value,mean_file类似于用caffe做图像分类时需要

//提供的lmdb文件，是关于每个像素点的均值，也就是对所有图片关于像素点

//均值， mean_value只有三个值，分别代表三个通道的均值，此函数初始化了

//均值图片mean_这一成员变量，此处不再详解
cv::Scalar channel_mean;
if (!mean_file.empty()) {
CHECK(mean_value.empty()) <<
"Cannot specify mean_file and mean_value at the same time";
BlobProto blob_proto;
ReadProtoFromBinaryFileOrDie(mean_file.c_str(), &blob_proto);

/* Convert from BlobProto to Blob<float> */
Blob<float> mean_blob;
mean_blob.FromProto(blob_proto);
CHECK_EQ(mean_blob.channels(), num_channels_)
<< "Number of channels of mean file doesn't match input layer.";

/* The format of the mean file is planar 32-bit float BGR or grayscale. */
std::vector<cv::Mat> channels;
float* data = mean_blob.mutable_cpu_data(); //data is a pointer that point the mean_blob
for (int i = 0; i < num_channels_; ++i) {
/* Extract an individual channel. */
cv::Mat channel(mean_blob.height(), mean_blob.width(), CV_32FC1, data); // 注意这种构造方式，将Mat的头指针与data同指向，而不进行拷贝，可去

4000
//docs.opencv.org见详细解释
channels.push_back(channel);
data += mean_blob.height() * mean_blob.width();
}

/* Merge the separate channels into a single image. */
cv::Mat mean;
cv::merge(channels, mean);

/* Compute the global mean pixel value and create a mean image
* filled with this value. */
channel_mean = cv::mean(mean);
mean_ = cv::Mat(input_geometry_, mean.type(), channel_mean);
}
if (!mean_value.empty()) {
CHECK(mean_file.empty()) <<
"Cannot specify mean_file and mean_value at the same time";
stringstream ss(mean_value);
vector<float> values;
string item;
while (getline(ss, item, ',')) {
float value = std::atof(item.c_str());
values.push_back(value);
}
CHECK(values.size() == 1 || values.size() == num_channels_) <<
"Specify either 1 mean_value or as many as channels: " << num_channels_;

std::vector<cv::Mat> channels;
for (int i = 0; i < num_channels_; ++i) {
/* Extract an individual channel. */
cv::Mat channel(input_geometry_.height, input_geometry_.width, CV_32FC1,
cv::Scalar(values[i]));                                            //利用scalar对Mat进行初始化的方式很好，

//例如Mat（height,width,CV_32F2,Scalar（1，2）），2个channel的图片，一层为1，

//一层为2
channels.push_back(channel);
}
cv::merge(channels, mean_);
}
}

/* Wrap the input layer of the network in separate cv::Mat objects
* (one per channel). This way we save one memcpy operation and we
* don't need to rely on cudaMemcpy2D. The last preprocessing
* operation will write the separate channels directly to the input //bang ding
* layer. */
void Detector::WrapInputLayer(std::vector<cv::Mat>* input_channels) {    //这里将输入input_channels与网络的输入绑定（wrap）

Blob<float>* input_layer = net_->input_blobs()[0];

int width = input_layer->width();
int height = input_layer->height();
float* input_data = input_layer->mutable_cpu_data();
for (int i = 0; i < input_layer->channels(); ++i) {
cv::Mat channel(height, width, CV_32FC1, input_data);               //就是这一步完成了绑定，将Mat的头指针与input_data指向相同，也就意味着，

//向Mat里写东西，就等同于向网络的输入写数据
input_channels->push_back(channel);
input_data += width * height;
}
}

void Detector::Preprocess(const cv::Mat& img,
std::vector<cv::Mat>* input_channels) {
/* Convert the input image to the input image format of the network. */              //各种初始化
cv::Mat sample;                                                                       //输入图片的channel与网络规定的channel不同，怎么办？
if (img.channels() == 3 && num_channels_ == 1)
cv::cvtColor(img, sample, cv::COLOR_BGR2GRAY);
else if (img.channels() == 4 && num_channels_ == 1)
cv::cvtColor(img, sample, cv::COLOR_BGRA2GRAY);
else if (img.channels() == 4 && num_channels_ == 3)
cv::cvtColor(img, sample, cv::COLOR_BGRA2BGR);
else if (img.channels() == 1 && num_channels_ == 3)
cv::cvtColor(img, sample, cv::COLOR_GRAY2BGR);
else
sample = img;

cv::Mat sample_resized;                                                               //输入图片的size，与网络规定不同，怎么办
if (sample.size() != input_geometry_)
cv::resize(sample, sample_resized, input_geometry_);
else
sample_resized = sample;

cv::Mat sample_float;                                                                //将像素值转化为float
if (num_channels_ == 3)
sample_resized.convertTo(sample_float, CV_32FC3);
else
sample_resized.convertTo(sample_float, CV_32FC1);

cv::Mat sample_normalized;
cv::subtract(sample_float, mean_, sample_normalized);                          //减去均值图像，0均值化

/* This operation will write the separate BGR planes directly to the
* input layer of the network because it is wrapped by the cv::Mat
* objects in input_channels. */
cv::split(sample_normalized, *input_channels);                             //这一步完成数据输入，感觉放在外面更容易理解，结合Detect方法看，上面提到

//input_channel已经和网络输入绑定，即指向相同，所以将数据写入input_channel的

//同时，就写入了网络输入

CHECK(reinterpret_cast<float*>(input_channels->at(0).data)
== net_->input_blobs()[0]->cpu_data())
<< "Input channels are not wrapping the input layer of the network.";
}
//一堆定义命令行输入的指令，配置Caffe时用到的一个包，好像是gflags实现的，第一个参数是名称，第二个是默认值，第三个是解释

//例如在命令行通过 -mean_value=" "就可以对mean_value进行赋值
DEFINE_string(mean_file, "",
"The mean file used to subtract from the input image.");
DEFINE_string(mean_value, "104,117,123",
"If specified, can be one value or can be same as image channels"
" - would subtract from the corresponding channel). Separated by ','."
"Either mean_file or mean_value should be provided, not both.");
DEFINE_string(file_type, "image",
"The file type in the list_file. Currently support image and video.");
DEFINE_string(out_file, "",
"If provided, store the detection results in the out_file.");
DEFINE_double(confidence_threshold, 0.01,
"Only store detections with score higher than the threshold.");
DEFINE_string(detect_type, "trace",
"Do detection:detect Do tracing :trace");

int main(int argc, char** argv) {
::google::InitGoogleLogging(argv[0]);
// Print output to stderr (while still logging)
FLAGS_alsologtostderr = 1;

#ifndef GFLAGS_GFLAGS_H_
namespace gflags = google;
#endif

gflags::SetUsageMessage("Do detection using SSD mode.\n"
"Usage:\n"
"    ssd_detect [FLAGS] model_file weights_file list_file\n");
gflags::ParseCommandLineFlags(&argc, &argv, true);

if (argc < 4) {
gflags::ShowUsageWithFlagsRestrict(argv[0], "examples/ssd/ssd_detect");
return 1;
}
//这里不再细讲，都是常规的操作
const string& model_file = argv[1];
const string& weights_file = argv[2];
const string& mean_file = FLAGS_mean_file;        //通过这种方式，读入上一步注释的赋值
const string& mean_value = FLAGS_mean_value;
const string& file_type = FLAGS_file_type;
const string& out_file = FLAGS_out_file;
const float confidence_threshold = FLAGS_confidence_threshold;
const string& detect_type=FLAGS_detect_type;

// Initialize the network.
Detector detector(model_file, weights_file, mean_file, mean_value);

// Set the output mode.
std::streambuf* buf = std::cout.rdbuf();
std::ofstream outfile;
if (!out_file.empty()) {
outfile.open(out_file.c_str());
if (outfile.good()) {
buf = outfile.rdbuf();
}
}
std::ostream out(buf);

// Process image one by one.
std::ifstream infile(argv[3]);
std::string file;
while (infile >> file) {
if (file_type == "image") {
cv::Mat img = cv::imread(file, -1);
CHECK(!img.empty()) << "Unable to decode image " << file;
std::vector<vector<float> > detections = detector.Detect(img);

/* Print the detection results. */
for (int i = 0; i < detections.size(); ++i) {
const vector<float>& d = detections[i];
// Detection format: [image_id, label, score, xmin, ymin, xmax, ymax].这里指明了那七个值
CHECK_EQ(d.size(), 7);
const float score = d[2];   //第三个是confidence
if (score >= confidence_threshold) {
out << file << " ";
out << static_cast<int>(d[1]) << " ";
out << score << " ";
out << static_cast<int>(d[3] * img.cols) << " ";
out << static_cast<int>(d[4] * img.rows) << " ";
out << static_cast<int>(d[5] * img.cols) << " ";
out << static_cast<int>(d[6] * img.rows) << std::endl;
}
}
} else if (file_type == "video" && detect_type=="trace") {
cv::VideoCapture cap(file);
if (!cap.isOpened()) {
LOG(FATAL) << "Failed to open video: " << file;
}
cv::Mat img;
int frame_count = 0;
while (true) {
bool success = cap.read(img);
if (!success) {
LOG(INFO) << "Process " << frame_count << " frames from " << file;
break;
}
CHECK(!img.empty()) << "Error when read frame";
std::vector<vector<float> > detections = detector.Detect(img);

/* Print the detection results. */
for (int i = 0; i < detections.size(); ++i) {
const vector<float>& d = detections[i];
// Detection format: [image_id, label, score, xmin, ymin, xmax, ymax].
CHECK_EQ(d.size(), 7);
const float score = d[2];
if (score >= confidence_threshold) {
out << file << "_";
out << std::setfill('0') << std::setw(6) << frame_count << " ";
out << static_cast<int>(d[1]) << " ";
out << score << " ";
out << static_cast<int>(d[3] * img.cols) << " ";
out << static_cast<int>(d[4] * img.rows) << " ";
out << static_cast<int>(d[5] * img.cols) << " ";
out << static_cast<int>(d[6] * img.rows) << std::endl;
}
}
++frame_count;
}
if (cap.isOpened()) {
cap.release();
}
} else if(file_type=="video" && detect_type=="detect"){
cv::VideoCapture cap(file);
bool detected=false;
if (!cap.isOpened()) {
LOG(FATAL) << "Failed to open video: " << file;
}
cv::Mat img;
int frame_count = 0;
while (true) {
bool success = cap.read(img);
if (!success) {
LOG(INFO) << "Process " << frame_count << " frames from " << file;
break;
}
CHECK(!img.empty()) << "Error when read frame";
std::vector<vector<float> > detections = detector.Detect(img);

/* Print the detection results. */
for (int i = 0; i < detections.size(); ++i) {
const vector<float>& d = detections[i];
// Detection format: [image_id, label, score, xmin, ymin, xmax, ymax].
CHECK_EQ(d.size(), 7);
const float score = d[2];
if (score >= confidence_threshold) {
out << file << "_";
out << std::setfill('0') << std::setw(6) << frame_count << " ";
out << static_cast<int>(d[1]) << " ";
out << score << " ";
out << static_cast<int>(d[3] * img.cols) << " ";
out << static_cast<int>(d[4] * img.rows) << " ";
out << static_cast<int>(d[5] * img.cols) << " ";
out << static_cast<int>(d[6] * img.rows) << std::endl;
detected=true;
break;
}
}
++frame_count;
if(detected==true) break;
}
if (cap.isOpened()) {
cap.release();
}
}else {
LOG(FATAL) << "Unknown file_type: " << file_type;
}
}
return 0;
}
#else
int main(int argc, char** argv) {
LOG(FATAL) << "This example requires OpenCV; compile with USE_OPENCV.";
}
#endif  // USE_OPENCV

至此，解释完了ssd_detect.cpp的内容，类似的可以看与之大体相同的cpp_classification的内容，也在examples下面，通过这个函数，就可以实现对于训练好的网络的调用，后期将会对如何利用C++对网络进行训练详解