https://blog.csdn.net/lovelyaiq/article/details/79929393

https://blog.csdn.net/qq_29462849/article/details/85272575

Opencv调用深度学习模型

2018年04月13日 15:19:54 TiRan_Yang 阅读数：1150更多

　　OpenCv 从V3.3版本开始支持调用深度学习模型,例如Caffe, Tensorflow, darknet等.详细见下图,具体的使用方法,可以参考官网:
https://docs.opencv.org/3.4.1/d6/d0f/group__dnn.html

　　目前Opencv可以支持的网络有GoogLeNet, ResNet-50,MobileNet-SSD from Caffe等,具体的可以参考:https://github.com/opencv/opencv/wiki/ChangeLog,里面有对dnn模块的详细介绍.
　　在github上,Opencv也有关于dnn模块的使用例子:https://github.com/opencv/opencv/tree/3.4.1/samples/dnn
　　这里只使用Python接口的Opencv 对Yolo V2(目前Opencv还不支持Yolo V3, 期待下一个版本支持)和Tensorflow训练出来的ssd_inception_v2_coco模型进行说明.

Yolo V2模型:

import cv2

import numpy as np

cap = cv2.VideoCapture('solidYellowLeft.mp4')

def read_cfg_model():

    model_path = '/home/scyang/TiRan/WorkSpace/others/darknet/cfg/yolov2.weights'

    cfg_path = '/home/scyang/TiRan/WorkSpace/others/darknet/cfg/yolov2.cfg'

    yolo_net = cv2.dnn.readNet(model_path, cfg_path, 'darknet')

    while True:

        flag, img = cap.read()

        if flag:

            yolo_net.setInput(cv2.dnn.blobFromImage(img, 1.0/127.5, (416, 416), (127.5, 127.5, 127.5), False, False))

            cvOut = yolo_net.forward()

            for detection in cvOut:

                confidence = np.max(detection[5:])

                if confidence > 0:

                    classIndex = np.argwhere(detection == confidence)[0][0] - 5

                    x_center = detection[0] * cols

                    y_center = detection[1] * rows

                    width = detection[2] * cols

                    height = detection[3] * rows

                    start = (int(x_center - width/2), int(y_center - height/2))

                    end = (int(x_center + width/2), int(y_center + height/2))

                    cv2.rectangle(img,start, end , (23, 230, 210), thickness=2)

        else:

            break

        cv2.imshow('show', img)

        cv2.waitKey(10)

　　这里需要对cvOut的结果说明一下：cvOut的前4个表示检测到的矩形框信息，第5位表示背景，从第6位开始代表检测到的目标置信度及目标属于那个类。
　　因此，下面两处的作用是，从5位开始获取结果中目标的置信度及目标属于那个类。

confidence = np.max(detection[5:])

classIndex = np.argwhere(detection == confidence)[0][0] - 5

结果的截图如下：

Tensorflow模型

cvNet = cv2.dnn.readNetFromTensorflow('model/ssd_inception_v2_coco_2017_11_17.pb','model/ssd_inception_v2_coco_2017_11_17.pbtxt')

while True:

    flag, img = cap.read()

    if flag:

        rows = img.shape[0]

        cols = img.shape[1]

        width = height = 300

        image = cv2.resize(img, ((int(cols * height / rows), width)))

        img = image[0:height, image.shape[1] - width:image.shape[1]]

        cvNet.setInput(cv2.dnn.blobFromImage(img, 1.0/127.5, (300, 300), (127.5, 127.5, 127.5), swapRB=True, crop=False))

        cvOut = cvNet.forward()

         # Network produces output blob with a shape 1x1xNx7 where N is a number of

         # detections and an every detection is a vector of values

         # [batchId, classId, confidence, left, top, right, bottom]

        for detection in cvOut[0,0,:,:]:

            score = float(detection[2])

            if score > 0.3:

                rows = cols = 300

                # print(detection)

                left = detection[3] * cols

                top = detection[4] * rows

                right = detection[5] * cols

                bottom = detection[6] * rows

                cv2.rectangle(img, (int(left), int(top)), (int(right), int(bottom)), (23, 230, 210), thickness=2)

        cv2.imshow('img', img)

        cv2.waitKey(10)

    else:

        break

效果如下：

　　使用方法和Yolo的类似，从最终的效果可以看出，ssd_inception_v2模型要比V2好。
注：blobFromImage的详细介绍及使用方法，可以参考某大神的博客：https://www.pyimagesearch.com/2017/11/06/deep-learning-opencvs-blobfromimage-works/。这里就不在多述了，要学会站在巨人的肩膀上

OpenCV4.0 Mask RCNN 实例分割示例 C＋/Python实现

文章来源：企鹅号 - 我爱计算机视觉

点击我爱计算机视觉标星，更快获取cvmL新技术

前几天OpenCV4.0-Alpha发布，其中新增实例分割Mask RCNN模型是这次发布的亮点之一。

图像实例分割即将图像中目标检测出来并进行像素级分割。

昨天learnopencv.com博主Satya Mallick发表博文，详述了使用新版OpenCV加载TensorFlow Object Detection Model Zone中的Mask RCNN模型实现目标检测与实例分割的应用。使用C++/Python实现的代码示例，都开源了。

先来看看作者发布的结果视频：

从视频可以看出，2.5GHZ i7 处理器每帧推断时间大约几百到2000毫秒。

TensorFlow Object Detection Model Zone中现在有四个使用不同骨干网（InceptionV2, ResNet50, ResNet101 和 Inception-ResnetV2）的Mask RCNN模型，这些模型都是在MSCOCO 数据库上训练出来的，其中使用Inception的模型是这四个中最快的。Satya Mallick博文中正是使用了该模型。

Mask RCNN网络架构

OpenCV使用Mask RCNN目标检测与实例分割流程：

1）下载模型。

地址：

http://download.tensorflow.org/models/object_detection/

现有的四个模型：

2）参数初始化。

设置目标检测的置信度阈值和Mask二值化分割阈值。

3）加载Mask RCNN模型、类名称与可视化颜色值。

mscoco_labels.names包含MSCOCO所有标注对象的类名称。

colors.txt是在图像上标出某实例时其所属类显示的颜色值。

frozen_inference_graph.pb模型权重。

mask_rcnn_inception_v2_coco_2018_01_28.pbtxt文本图文件，告诉OpenCV如何加载模型权重。

OpenCV已经给定工具可以从给定模型权重提取出文本图文件。详见：

https://github.com/opencv/opencv/wiki/TensorFlow-Object-Detection-API

OpenCV支持CPU和OpenCL推断，但OpenCL只支持Intel自家GPU，Satya设置了CPU推断模式（cv.dnn.DNN_TARGET_CPU）。

4）读取图像、视频或者摄像头数据。

5）对每一帧数据计算处理。

主要步骤如图：

6）提取目标包围框和Mask，并绘制结果。

C++/Python代码下载：

https://github.com/spmallick/learnopencv/tree/master/Mask-RCNN

原博文地址：

https://www.learnopencv.com/deep-learning-based-object-detection-and-instance-segmentation-using-mask-r-cnn-in-opencv-python-c/

【点赞与转发】就是一种鼓励

C++调用mask rcnn进行实时检测--opencv4.0

2018年12月27日 09:35:43 Oliver Cui 阅读数：52

介绍

Opencv在前面的几个版本中已经支持caffe、tensorflow、pytorch训练的几种模型，包括分类和物体检测模型（SSD、Yolo），针对tensorflow，opencv与tensorflow object detection api对接，可以通过该api训练模型，然后通过opencv调用，这样就可以把python下的环境移植到C++中。

关于tensorflow object detection api，后面博文会详细介绍

数据准备与环境配置

基于mask_rcnn_inception_v2_coco_2018_01_28的frozen_inference_graph.pb，这个模型在tensorflow object detection api中可以找到，然后需要对应的mask_rcnn_inception_v2_coco_2018_01_28.pbtxt，以及colors.txt，mscoco_labels.names。

opencv必须是刚发布的4.0版本，该版本支持mask rcnn和faster rcnn，低版本不支持哦，注意opencv4.0中在配置环境时，include下少了一个opencv文件夹，只有opencv2，这是正常的。

好了，废话不多说了，直接上源代码，该代码调用usb摄像头进行实时检测，基于单幅图像的检测修改下代码即可。



#include <fstream>

#include <sstream>

#include <iostream>

#include <string.h>

#include <opencv2/dnn.hpp>

#include <opencv2/imgproc.hpp>

#include <opencv2/highgui.hpp>

using namespace cv;

using namespace dnn;

using namespace std;

// Initialize the parameters

float confThreshold = 0.5; // Confidence threshold

float maskThreshold = 0.3; // Mask threshold

vector<string> classes;

vector<Scalar> colors;

// Draw the predicted bounding box

void drawBox(Mat& frame, int classId, float conf, Rect box, Mat& objectMask);

// Postprocess the neural network's output for each frame

void postprocess(Mat& frame, const vector<Mat>& outs);

int main()

{

	// Load names of classes

	string classesFile = "./mask_rcnn_inception_v2_coco_2018_01_28/mscoco_labels.names";

	ifstream ifs(classesFile.c_str());

	string line;

	while (getline(ifs, line)) classes.push_back(line);

	// Load the colors

	string colorsFile = "./mask_rcnn_inception_v2_coco_2018_01_28/colors.txt";

	ifstream colorFptr(colorsFile.c_str());

	while (getline(colorFptr, line))

	{

		char* pEnd;

		double r, g, b;

		r = strtod(line.c_str(), &pEnd);

		g = strtod(pEnd, NULL);

		b = strtod(pEnd, NULL);

		Scalar color = Scalar(r, g, b, 255.0);

		colors.push_back(Scalar(r, g, b, 255.0));

	}

	// Give the configuration and weight files for the model

	String textGraph = "./mask_rcnn_inception_v2_coco_2018_01_28/mask_rcnn_inception_v2_coco_2018_01_28.pbtxt";

	String modelWeights = "./mask_rcnn_inception_v2_coco_2018_01_28/frozen_inference_graph.pb";

	// Load the network

	Net net = readNetFromTensorflow(modelWeights, textGraph);

	net.setPreferableBackend(DNN_BACKEND_OPENCV);

	net.setPreferableTarget(DNN_TARGET_CPU);

	// Open a video file or an image file or a camera stream.

	string str, outputFile;

	VideoCapture cap(0);//根据摄像头端口id不同，修改下即可

	//VideoWriter video;

	Mat frame, blob;

	// Create a window

	static const string kWinName = "Deep learning object detection in OpenCV";

	namedWindow(kWinName, WINDOW_NORMAL);

	// Process frames.

	while (waitKey(1) < 0)

	{

		// get frame from the video

		cap >> frame;

		// Stop the program if reached end of video

		if (frame.empty())

		{

			cout << "Done processing !!!" << endl;

			cout << "Output file is stored as " << outputFile << endl;

			waitKey(3000);

			break;

		}

		// Create a 4D blob from a frame.

		blobFromImage(frame, blob, 1.0, Size(frame.cols, frame.rows), Scalar(), true, false);

		//blobFromImage(frame, blob);

		//Sets the input to the network

		net.setInput(blob);

		// Runs the forward pass to get output from the output layers

		std::vector<String> outNames(2);

		outNames[0] = "detection_out_final";

		outNames[1] = "detection_masks";

		vector<Mat> outs;

		net.forward(outs, outNames);

		// Extract the bounding box and mask for each of the detected objects

		postprocess(frame, outs);

		// Put efficiency information. The function getPerfProfile returns the overall time for inference(t) and the timings for each of the layers(in layersTimes)

		vector<double> layersTimes;

		double freq = getTickFrequency() / 1000;

		double t = net.getPerfProfile(layersTimes) / freq;

		string label = format("Mask-RCNN on 2.5 GHz Intel Core i7 CPU, Inference time for a frame : %0.0f ms", t);

		putText(frame, label, Point(0, 15), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 0));

		// Write the frame with the detection boxes

		Mat detectedFrame;

		frame.convertTo(detectedFrame, CV_8U);

		imshow(kWinName, frame);

	}

	cap.release();

	return 0;

}

// For each frame, extract the bounding box and mask for each detected object

void postprocess(Mat& frame, const vector<Mat>& outs)

{

	Mat outDetections = outs[0];

	Mat outMasks = outs[1];

	// Output size of masks is NxCxHxW where

	// N - number of detected boxes

	// C - number of classes (excluding background)

	// HxW - segmentation shape

	const int numDetections = outDetections.size[2];

	const int numClasses = outMasks.size[1];

	outDetections = outDetections.reshape(1, outDetections.total() / 7);

	for (int i = 0; i < numDetections; ++i)

	{

		float score = outDetections.at<float>(i, 2);

		if (score > confThreshold)

		{

			// Extract the bounding box

			int classId = static_cast<int>(outDetections.at<float>(i, 1));

			int left = static_cast<int>(frame.cols * outDetections.at<float>(i, 3));

			int top = static_cast<int>(frame.rows * outDetections.at<float>(i, 4));

			int right = static_cast<int>(frame.cols * outDetections.at<float>(i, 5));

			int bottom = static_cast<int>(frame.rows * outDetections.at<float>(i, 6));

			left = max(0, min(left, frame.cols - 1));

			top = max(0, min(top, frame.rows - 1));

			right = max(0, min(right, frame.cols - 1));

			bottom = max(0, min(bottom, frame.rows - 1));

			Rect box = Rect(left, top, right - left + 1, bottom - top + 1);

			// Extract the mask for the object

			Mat objectMask(outMasks.size[2], outMasks.size[3], CV_32F, outMasks.ptr<float>(i, classId));

			// Draw bounding box, colorize and show the mask on the image

			drawBox(frame, classId, score, box, objectMask);

		}

	}

}

// Draw the predicted bounding box, colorize and show the mask on the image

void drawBox(Mat& frame, int classId, float conf, Rect box, Mat& objectMask)

{

	//Draw a rectangle displaying the bounding box

	rectangle(frame, Point(box.x, box.y), Point(box.x + box.width, box.y + box.height), Scalar(255, 178, 50), 3);

	//Get the label for the class name and its confidence

	string label = format("%.2f", conf);

	if (!classes.empty())

	{

		CV_Assert(classId < (int)classes.size());

		label = classes[classId] + ":" + label;

	}

	//Display the label at the top of the bounding box

	int baseLine;

	Size labelSize = getTextSize(label, FONT_HERSHEY_SIMPLEX, 0.5, 1, &baseLine);

	box.y = max(box.y, labelSize.height);

	rectangle(frame, Point(box.x, box.y - round(1.5*labelSize.height)), Point(box.x + round(1.5*labelSize.width), box.y + baseLine), Scalar(255, 255, 255), FILLED);

	putText(frame, label, Point(box.x, box.y), FONT_HERSHEY_SIMPLEX, 0.75, Scalar(0, 0, 0), 1);

	Scalar color = colors[classId%colors.size()];

	// Resize the mask, threshold, color and apply it on the image

	resize(objectMask, objectMask, Size(box.width, box.height));

	Mat mask = (objectMask > maskThreshold);

	Mat coloredRoi = (0.3 * color + 0.7 * frame(box));

	coloredRoi.convertTo(coloredRoi, CV_8UC3);

	// Draw the contours on the image

	vector<Mat> contours;

	Mat hierarchy;

	mask.convertTo(mask, CV_8U);

	findContours(mask, contours, hierarchy, RETR_CCOMP, CHAIN_APPROX_SIMPLE);

	drawContours(coloredRoi, contours, -1, color, 5, LINE_8, hierarchy, 100);

	coloredRoi.copyTo(frame(box), mask);

}

实验结果

不过检测速度很慢，I7-8700k，GTX1060下需要1s每帧，达不到实时性要求。。。

实验数据

本博文所有的数据可以从这里下载：opencv调用mask rcnn数据