opencv进阶-SSD模型实时人脸检测

reference：利用OpenCV和深度学习实现人脸检测
2018年之后，opencv DNN模式下开始使用卷积神经网络SSD人脸检测器，目前商业应用非常成熟，可以做到实时运行，对各种角度人脸均能做到准确的检测，具有很强的抗干扰性。

opencv自带的人脸检测模型

一、先安装opencv

二、下载模型文件

打开windows下的终端，点击左下角的徽标键，输入cmd即可。然后在终端输入cd /d D:\opencv-4.1.0\opencv\sources\samples\dnn\face_detector
然后继续输入：python download_weights.py，这样就会根据权重文件生成配置文件和模型文件。

三、查看参数

进入models.yml文件，查看需要设置的参数。文件路径：D:\opencv-4.1.0\opencv\sources\samples\dnn

四、全部代码

1.基于caffe

基于caffe框架下的SSD深度卷积神经网络模型，做人脸检测。
.prototxt和.caffemodel的作用如下：

The .prototxt file(s) which define the model architecture (i.e., the layers themselves)

The .caffemodel file which contains the weights for the actual layers

#include 
#include 
#include 

using namespace cv;
using namespace cv::dnn;
using namespace std;

String model_text_file = "D:/opencv-4.1.0/models/face_detector/deploy.prototxt";
String modelFile = "D:/opencv-4.1.0/models/face_detector/res10_300x300_ssd_iter_140000_fp16.caffemodel";

int main(int argc, char** argv) {
	VideoCapture capture;
	capture.open(0);
	namedWindow("input", WINDOW_AUTOSIZE);
	int w = capture.get(CAP_PROP_FRAME_WIDTH);
	int h = capture.get(CAP_PROP_FRAME_HEIGHT);
	printf("frame width : %d, frame height : %d", w, h);

	// set up net
	Net net = readNetFromCaffe(model_text_file, modelFile);
	net.setPreferableBackend(DNN_BACKEND_OPENCV);
	net.setPreferableTarget(DNN_TARGET_CPU);

	Mat frame;
	while (capture.read(frame)) {
		flip(frame, frame, 1);
		imshow("input", frame);

		//预测
		Mat inputblob = blobFromImage(frame, 1.0, Size(300, 300), Scalar(104,177,123), false);
		net.setInput(inputblob, "data");
		Mat detection = net.forward("detection_out");

		//检测
		Mat detectionMat(detection.size[2], detection.size[3], CV_32F, detection.ptr<float>());
		float confidence_threshold = 0.25;
		for (int i = 0; i < detectionMat.rows; i++) {
			float confidence = detectionMat.at<float>(i, 2);
			if (confidence > confidence_threshold) {
				size_t objIndex = (size_t)(detectionMat.at<float>(i, 1));
				float tl_x = detectionMat.at<float>(i, 3) * frame.cols;
				float tl_y = detectionMat.at<float>(i, 4) * frame.rows;
				float br_x = detectionMat.at<float>(i, 5) * frame.cols;
				float br_y = detectionMat.at<float>(i, 6) * frame.rows;

				Rect object_box((int)tl_x, (int)tl_y, (int)(br_x - tl_x), (int)(br_y - tl_y));
				rectangle(frame, object_box, Scalar(0, 0, 255), 2, 8, 0);
				putText(frame, format("%s %.2f", "face", confidence), Point(tl_x, tl_y), FONT_HERSHEY_SIMPLEX, 1.0, Scalar(255, 0, 0), 2);
			}
		}
		vector < double>layerstimings;
		double freq = getTickFrequency() / 1000;
		double time = net.getPerfProfile(layerstimings) / freq;
		ostringstream ss;
		ss << "FPS" << 1000 / time << ";time:" << time << "ms";
		putText(frame, ss.str(), Point(20, 20), FONT_HERSHEY_PLAIN, 1, Scalar(0, 0, 255), 2, 8);
		imshow("face-detection-demo", frame);
		char c = waitKey(5);
		if (c == 27) { // ESC退出
			break;
		}
	}
	capture.release();//释放资源
	waitKey(0);
	return 0;
}

2.基于TensorFlow

基于TensorFlow的SSD模型，人脸检测。

#include 
#include 
#include 

using namespace cv;
using namespace cv::dnn;
using namespace std;

String model_text_file = "D:/opencv-4.1.0/models/face_detector/opencv_face_detector.pbtxt";
String modelFile = "D:/opencv-4.1.0/models/face_detector/opencv_face_detector_uint8.pb";

int main(int argc, char** argv) {
	VideoCapture capture;
	capture.open(0);
	namedWindow("input", WINDOW_AUTOSIZE);
	int w = capture.get(CAP_PROP_FRAME_WIDTH);
	int h = capture.get(CAP_PROP_FRAME_HEIGHT);
	printf("frame width : %d, frame height : %d", w, h);

	// set up net
	Net net = readNetFromTensorflow(modelFile, model_text_file);

	net.setPreferableBackend(DNN_BACKEND_OPENCV);
	net.setPreferableTarget(DNN_TARGET_CPU);

	Mat frame;
	while (capture.read(frame)) {
		flip(frame, frame, 1);
		imshow("input", frame);

		//预测
		Mat inputblob = blobFromImage(frame, 1.0, Size(300, 300), Scalar(104,177,123), false);
		net.setInput(inputblob, "data");
		Mat detection = net.forward("detection_out");

		//检测
		Mat detectionMat(detection.size[2], detection.size[3], CV_32F, detection.ptr<float>());
		float confidence_threshold = 0.25;
		for (int i = 0; i < detectionMat.rows; i++) {
			float confidence = detectionMat.at<float>(i, 2);
			if (confidence > confidence_threshold) {
				size_t objIndex = (size_t)(detectionMat.at<float>(i, 1));
				float tl_x = detectionMat.at<float>(i, 3) * frame.cols;
				float tl_y = detectionMat.at<float>(i, 4) * frame.rows;
				float br_x = detectionMat.at<float>(i, 5) * frame.cols;
				float br_y = detectionMat.at<float>(i, 6) * frame.rows;

				Rect object_box((int)tl_x, (int)tl_y, (int)(br_x - tl_x), (int)(br_y - tl_y));
				rectangle(frame, object_box, Scalar(0, 0, 255), 2, 8, 0);
				putText(frame, format("%s %.2f", "face", confidence), Point(tl_x, tl_y), FONT_HERSHEY_SIMPLEX, 1.0, Scalar(255, 0, 0), 2);
			}
		}
		vector < double>layerstimings;
		double freq = getTickFrequency() / 1000;
		double time = net.getPerfProfile(layerstimings) / freq;
		ostringstream ss;
		ss << "FPS" << 1000 / time << ";time:" << time << "ms";
		putText(frame, ss.str(), Point(20, 20), FONT_HERSHEY_PLAIN, 1, Scalar(0, 0, 255), 2, 8);
		imshow("face-detection-demo", frame);
		char c = waitKey(5);
		if (c == 27) { // ESC退出
			break;
		}
	}
	capture.release();//释放资源
	waitKey(0);
	return 0;
}

在笔记本上运行，还挺流程的。