OpenCV如何創建一個DescriptorExtractor對象

Question

我正在使用OpenCV C++庫，但我沒有設法創建一個DescriptorExtractor對象。 這裏是我做過什麼：

Mat img = imread("testOrb.jpg",CV_LOAD_IMAGE_UNCHANGED); 
std::vector<KeyPoint> kp; 
cv::Ptr<cv::ORB> detector = cv::ORB::create(); 
detector->detect(img, kp) 
//this part works  

DescriptorExtractor descriptorExtractor;  
Mat descriptors; 
descriptorExtractor.compute(img, kp, descriptors); 
//when these 3 lines are added, an error is thrown 

但我有以下錯誤信息：

OpenCV Error: The function/feature is not implemented() in detectAndCompute, file ... 

Answer 1

DescriptorExtractor是抽象類，所以你不能實例化。這只是描述符提取器的通用接口。你可以這樣做：

Ptr<DescriptorExtractor> descriptorExtractor = ORB::create(); 
Mat descriptors; 
descriptorExtractor->compute(img, kp, descriptors); 

注意，也存在FeatureDetector，這是關鍵點檢測的通用接口，這樣你就可以這樣做：

std::vector<KeyPoint> kp; 
Ptr<FeatureDetector> detector = ORB::create(); 
detector->detect(img, kp);