ImageRecognition/IDetection.h

#pragma once
#include<iostream>
#include <vector>
#include<opencv2/opencv.hpp>
#include "YunDaISASImageRecognitionService.h"
using namespace std;
class IDetection {
public:
	struct Output {
		Output(int id, float confidence, cv::Rect box)
		{
			m_id = id;
			m_confidence = confidence;
			m_box = box;
		}
		/*Output() 
		{

		}*/
	public:
		int m_id;             //结果类别id
		float m_confidence;   //结果置信度
		cv::Rect m_box;       //矩形框
	};

public:
	struct DectectResult {
	//	DectectResult(float confidence, float dValue, string sValue/*, string sUnit,string sName*/ )
	//	{
	//		m_confidence = confidence;
	//		m_dValue = dValue;
	//		m_sValue = sValue;
	///*		m_sUnit = sUnit;
	//		m_sName = sName;*/
	//	}
		DectectResult(float confidence, float dValue, string sValue)
		{
			m_confidence = confidence;
			m_dValue = dValue;
			m_sValue = sValue;
		}
		DectectResult()
		{
		}
		float m_confidence =0;   //结果置信度
		float m_dValue =0;
		string m_sValue="";
		//string m_sUnit=""; //单位 例如℃
		//string m_sName=""; //名称


	};
//public:
//    virtual vector<DectectResult> GetStateResult(cv::Mat img,std::vector<cv::Rect> rects) = 0;
//    virtual vector<DectectResult> GetDigitResult(cv::Mat img,std::vector<cv::Rect> rects) = 0;
public:
#if(defined YOLO_P6 && YOLO_P6==true)
	const float netAnchors[4][6] = { { 19,27, 44,40, 38,94 },{ 96,68, 86,152, 180,137 },{ 140,301, 303,264, 238,542 },{ 436,615, 739,380, 925,792 } };

	const int netWidth = 1280;  //ONNX图片输入宽度
	const int netHeight = 1280; //ONNX图片输入高度

	const int strideSize = 4;  //stride size
#else
	const float netAnchors[3][6] = { { 10,13, 16,30, 33,23 },{ 30,61, 62,45, 59,119 },{ 116,90, 156,198, 373,326 } };

	const int netWidth = 640;   //ONNX图片输入宽度
	const int netHeight = 640;  //ONNX图片输入高度

	const int strideSize = 3;   //stride size
#endif // YOLO_P6

	const float netStride[4] = { 8, 16.0,32,64 };

	float boxThreshold = 0.25;
	float classThreshold = 0.25;

	float nmsThreshold = 0.45;
	float nmsScoreThreshold = boxThreshold * classThreshold;
public:
	static void DrawPred(cv::Mat& img, vector<Output> result,vector<std::string> className)
	{
		for (int i = 0; i < result.size(); i++) {
			try
			{
				int left = result[i].m_box.x;
				int top = result[i].m_box.y;
				int width = result[i].m_box.width;
				int height = result[i].m_box.height;
				int baseLine;
				//2022.10.17
				cv::Rect box = result[i].m_box;
				cv::Mat input_image_copy = img.clone();
				cv::Mat cutMat = input_image_copy(box);
				string label = className[result[i].m_id] + ": " + to_string(result[i].m_confidence);
				cv::Size labelSize = getTextSize(label, cv::FONT_HERSHEY_SIMPLEX, 0.5, 1, &baseLine);
				top = max(top, labelSize.height);
				cv::rectangle(img, cv::Point(left, top - int(2 * labelSize.height)), cv::Point(left + int(2 * labelSize.width), top + baseLine), cv::Scalar(0, 0, 255), cv::FILLED);
				cv::rectangle(img, cv::Point(left, top), cv::Point(left + width, top + height), cv::Scalar(255, 255, 0), 1);
				cv::putText(img, label, cv::Point(left, top), cv::FONT_HERSHEY_SIMPLEX, 1, cv::Scalar(255, 255, 0), 1);
			
			}
			catch (const std::exception& ex)
			{
				YunDaISASImageRecognitionService::ConsoleLog(ex.what());
			}
		}
	}

	
};