Problem building a ROS package with OpenCV 2.4.9

Question

I am trying to build a ROS package, using catkin build. The package was made to run with OpenCV2, so I installed OpenCV 2.4.9 from source.

I faced some errors when running catkin build. The situation is as following:

At first, I changed the CMakeLists.txt, so that it can locate the right version of OpenCV.
Then, I added #include <vector> and changed every occurence of vector to std::vector in the opencv2/nonfree/features2d.hpp file, since I was getting the error that vector was not declared and the error disappeared.

Another error I get is:

In file included from /usr/local/include/opencv2/nonfree/nonfree.hpp:46:0,
from /home/vm/catkin_ws/src/ogm_merging/ogm_evaluation/feature_evaluation/include/feature_evaluation/feature_evaluation_metrics/feature_metrics.h:29:
from /home/vm/catkin_ws/src/ogm_merging/ogm_evaluation/feature_evaluation/src/feature_evaluation/feature_evaluation_metrics/feature_metrics.cpp:19:
 /usr/local/include/opencv2/nonfree/features2d.hpp:133:5:   
 error: ‘AlgorithmInfo’ does not name a type
      AlgorithmInfo* info() const;
      ^

The nonfree/features2d.hpp is as following:

#ifndef __OPENCV_NONFREE_FEATURES_2D_HPP__
#define __OPENCV_NONFREE_FEATURES_2D_HPP__

#include "opencv2/features2d/features2d.hpp"

#ifdef __cplusplus

namespace cv
{

/*!
 SIFT implementation.
 The class implements SIFT algorithm by D. Lowe.
*/
class CV_EXPORTS_W SIFT : public Feature2D
{
public:

    CV_WRAP explicit SIFT( int nfeatures=0, int nOctaveLayers=3,
          double contrastThreshold=0.04, double edgeThreshold=10,
          double sigma=1.6);

    //! returns the descriptor size in floats (128)
    CV_WRAP int descriptorSize() const;

    //! returns the descriptor type
    CV_WRAP int descriptorType() const;

    //! finds the keypoints using SIFT algorithm
    void operator()(InputArray img, InputArray mask,
                    std::vector<KeyPoint>& keypoints) const;
    //! finds the keypoints and computes descriptors for them using SIFT algorithm.
    //! Optionally it can compute descriptors for the user-provided keypoints
    void operator()(InputArray img, InputArray mask,
                    std::vector<KeyPoint>& keypoints,
                    OutputArray descriptors,
                    bool useProvidedKeypoints=false) const;

    AlgorithmInfo* info() const;

    void buildGaussianPyramid( const Mat& base, std::vector<Mat>& pyr, int nOctaves ) const;

    void buildDoGPyramid( const std::vector<Mat>& pyr, std::vector<Mat>& dogpyr ) const;

    void findScaleSpaceExtrema( const std::vector<Mat>& gauss_pyr, const std::vector<Mat>& dog_pyr,
                                std::vector<KeyPoint>& keypoints ) const;
protected:
    void detectImpl( const Mat& image, std::vector<KeyPoint>& keypoints, const Mat& mask=Mat() ) const;
    void computeImpl( const Mat& image, std::vector<KeyPoint>& keypoints, Mat& descriptors ) const;

    CV_PROP_RW int nfeatures;
    CV_PROP_RW int nOctaveLayers;
    CV_PROP_RW double contrastThreshold;
    CV_PROP_RW double edgeThreshold;
    CV_PROP_RW double sigma;
};

typedef SIFT SiftFeatureDetector;
typedef SIFT SiftDescriptorExtractor;

} /* namespace cv */
#endif /* __cplusplus */
#endif

I do not want to change to OpenCV 3 and add the extra modules, as many people in here say, I want to make this run with OpenCV2.

I tried to include files, such as core/core.hpp of features2d/features2d.hpp, but nothing changes. Where is the AlgorithmInfo class defined, so that I can use it?

I am using Ubuntu 16.04 and ROS Kinetic.

Update: It worked with Ubuntu 14.04, ROS Indigo and Opencv 2.4, so the problem must be related to ROS or Ubuntu version. Any ideas about it?

Thank you in advance.

Answer 1

ROS and OpenCV works together through opencv_bridge. I am not sure if this will fix your initial problem, but it looks like you are not using opencv_bridge which will solve at least some of your problems.