The DNN of OpenCV is used to read onnx as the prediction end to realize classification

Be careful ：
1、size Consistent with training ;
2、 Used in training Normalize Words , because blobFromImage Variance is not supported , Need to rewrite blobFromImage;
3、 Format of prediction results , Ensure that the corresponding indicators are read correctly ;

class ModelFlipClass
{
    
private:
	Net net;
public:
	bool Read(string pathNet);
	int Detect(string strImgPath);
};

bool ModelFlipClass::Read(string pathNet)
{
    
	try {
    
		net = readNet(pathNet);
	}
	catch (const std::exception&) {
    
		return false;
	}
	net.setPreferableBackend(cv::dnn::DNN_BACKEND_DEFAULT);
	net.setPreferableTarget(cv::dnn::DNN_TARGET_CPU);
	return true;
}

int ModelXRayFlipClass::Detect(string strImgPath)
{
    
	try 
	{
    
		Mat imgInput = imread(strImgPath, IMREAD_UNCHANGED);
		Mat imgNorm;
		normalize(imgInput, imgNorm, 0, 255, NORM_MINMAX);//16 The bit changes to 8 position 
		convertScaleAbs(imgNorm, imgNorm);
		Mat imRGB;
		cvtColor(imgNorm, imRGB, COLOR_GRAY2BGR);
		const int nHight = 400;// When training is needed transforms Of Resize bring into correspondence with 
		const int nWight = 200;
		Mat blob;
		blobFromImage(imRGB, blob, 1 / 255.0, cv::Size(nWight, nHight), Scalar(), true, false);
		net.setInput(blob);
		vector<cv::Mat> netOutputImg;
		net.forward(netOutputImg, net.getUnconnectedOutLayersNames());
		float* pdata = (float*)netOutputImg[0].data;
		vector<float> vecScore;
		for (size_t i = 0; i < 2; i++) // There are two kinds of , So here is 2
		{
    
			vecScore.push_back(pdata[i]);
		}
		auto maxPosition = max_element(vecScore.begin(), vecScore.end());
		int nMaxIndex = maxPosition - vecScore.begin();//vdbAmplitude The index when the absolute value of the value in is the largest 
		return nMaxIndex;
	}
	catch (const std::exception&) {
    
		return -1;
	}
}