Extracción de puntos clave

Usaremos los algoritmos ORB, BRISK, AKAZE, entre otros.

Author: devcarmelo

opencv-keypoints/CMakeLists.txt

@@ -0,0 +1,10 @@
+cmake_minimum_required(VERSION 3.4)
+
+project( Tutorial_Keypoints )
+
+find_package( OpenCV 3.0.0 REQUIRED )
+
+file(COPY images DESTINATION data)
+
+add_executable( ${PROJECT_NAME} source.cpp)
+target_link_libraries( ${PROJECT_NAME} ${OpenCV_LIBS} )

opencv-keypoints/images/box.png

@@ -0,0 +1,47 @@
+#include <iostream>
+#include <opencv2\opencv.hpp>
+#include <opencv2\features2d.hpp>
+
+using namespace cv;
+using namespace std;
+
+int main(int argc, char** argv )
+{
+    Mat imageA = imread("data/images/box.png", IMREAD_GRAYSCALE);
+    Mat imageB = imread("data/images/box_in_scene.png", IMREAD_GRAYSCALE);
+
+    if (imageA.empty() || imageB.empty())
+    {
+        printf("No image data.");
+        getchar();
+
+        return -1;
+    }
+
+    Ptr<Feature2D> detect = BRISK::create();
+
+    vector<KeyPoint> kpA, kpB;
+    Mat descA, descB;
+
+    detect->detectAndCompute(imageA, noArray(), kpA, descA);
+    detect->detectAndCompute(imageB, noArray(), kpB, descB);
+
+    vector<DMatch> matches;
+
+    Ptr<DescriptorMatcher> matcher = BFMatcher::create(NORM_HAMMING, true);
+    matcher->match(descA, descB, matches);
+
+    sort(matches.begin(), matches.end());
+    matches.erase(matches.begin() + 35, matches.end());
+
+    Scalar color = Scalar::all(-1);
+    int flags = DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS;
+
+    Mat result;
+    drawMatches(imageA, kpA, imageB, kpB, matches, result, color, color, vector<char>(), flags);
+
+    imshow("OpenCV :: Match Keypoints", result);
+    waitKey(0);
+
+    return 0;
+}