CFeatureExtraction_FAST.cpp

Go to the documentation of this file.

/* +------------------------------------------------------------------------+
   |                     Mobile Robot Programming Toolkit (MRPT)            |
   |                          http://www.mrpt.org/                          |
   |                                                                        |
   | Copyright (c) 2005-2017, Individual contributors, see AUTHORS file     |
   | See: http://www.mrpt.org/Authors - All rights reserved.                |
   | Released under BSD License. See details in http://www.mrpt.org/License |
   +------------------------------------------------------------------------+ */

#include "vision-precomp.h"  // Precompiled headers

#include <mrpt/vision/CFeatureExtraction.h>

// Universal include for all versions of OpenCV
#include <mrpt/otherlibs/do_opencv_includes.h>

using namespace mrpt;
using namespace mrpt::vision;
using namespace mrpt::system;
using namespace mrpt::utils;
using namespace mrpt::math;
using namespace std;

/************************************************************************************************
*                                                               extractFeaturesFAST
**
************************************************************************************************/
void CFeatureExtraction::extractFeaturesFAST(
        const mrpt::utils::CImage& inImg, CFeatureList& feats, unsigned int init_ID,
        unsigned int nDesiredFeatures, const TImageROI& ROI,
        const CMatrixBool* mask) const
{
        MRPT_UNUSED_PARAM(ROI);
        MRPT_START

#if MRPT_HAS_OPENCV
#if MRPT_OPENCV_VERSION_NUM < 0x210
        THROW_EXCEPTION("This function requires OpenCV > 2.1.0")
#else

        using namespace cv;

        vector<KeyPoint> cv_feats;  // The opencv keypoint output vector

        // Make sure we operate on a gray-scale version of the image:
        const CImage inImg_gray(inImg, FAST_REF_OR_CONVERT_TO_GRAY);

// JL: Instead of
//      int aux = options.FASTOptions.threshold; ....
//  It's better to use an adaptive threshold, controlled from our caller
//  outside.

#if MRPT_OPENCV_VERSION_NUM >= 0x211

        //    cv::Mat *mask ;
        //    if( _mask )
        //       mask = static_cast<cv::Mat*>(_mask);

        const Mat theImg = cvarrToMat(inImg_gray.getAs<IplImage>());

        cv::Mat cvMask;
        if (options.useMask)
        {
                cout << "using mask" << endl;
                size_t maskW = mask->getColCount(), maskH = mask->getRowCount();
                ASSERT_(
                        maskW == inImg_gray.getWidth() && maskH == inImg_gray.getHeight());

                // Convert Mask into CV type
                cvMask = cv::Mat::ones(maskH, maskW, CV_8UC1);
                for (int ii = 0; ii < int(maskW); ++ii)
                        for (int jj = 0; jj < int(maskH); ++jj)
                        {
                                if (!mask->get_unsafe(jj, ii))
                                {
                                        cvMask.at<char>(ii, jj) = (char)0;
                                }
                        }
        }

#if MRPT_OPENCV_VERSION_NUM < 0x300
        FastFeatureDetector fastDetector(
                options.FASTOptions.threshold, options.FASTOptions.nonmax_suppression);
        fastDetector.detect(theImg, cv_feats);
#else
        Ptr<cv::FastFeatureDetector> fastDetector = cv::FastFeatureDetector::create(
                options.FASTOptions.threshold, options.FASTOptions.nonmax_suppression);
        fastDetector->detect(theImg, cv_feats);
#endif

#elif MRPT_OPENCV_VERSION_NUM >= 0x210
        FAST(
                inImg_gray.getAs<IplImage>(), cv_feats, options.FASTOptions.threshold,
                options.FASTOptions.nonmax_suppression);
#endif

        // *All* the features have been extracted.
        const size_t N = cv_feats.size();

        // Use KLT response instead of the OpenCV's original "response" field:
        if (options.FASTOptions.use_KLT_response)
        {
                const unsigned int KLT_half_win = 4;
                const unsigned int max_x = inImg_gray.getWidth() - 1 - KLT_half_win;
                const unsigned int max_y = inImg_gray.getHeight() - 1 - KLT_half_win;
                for (size_t i = 0; i < N; i++)
                {
                        const unsigned int x = cv_feats[i].pt.x;
                        const unsigned int y = cv_feats[i].pt.y;
                        if (x > KLT_half_win && y > KLT_half_win && x <= max_x &&
                                y <= max_y)
                                cv_feats[i].response =
                                        inImg_gray.KLT_response(x, y, KLT_half_win);
                        else
                                cv_feats[i].response = -100;
                }
        }

        // Now:
        //  1) Sort them by "response": It's ~100 times faster to sort a list of
        //      indices "sorted_indices" than sorting directly the actual list of
        //      features "cv_feats"
        std::vector<size_t> sorted_indices(N);
        for (size_t i = 0; i < N; i++) sorted_indices[i] = i;
        std::sort(
                sorted_indices.begin(), sorted_indices.end(),
                KeypointResponseSorter<vector<KeyPoint>>(cv_feats));

        //  2) Filter by "min-distance" (in options.FASTOptions.min_distance)
        //  3) Convert to MRPT CFeatureList format.
        // Steps 2 & 3 are done together in the while() below.
        // The "min-distance" filter is done by means of a 2D binary matrix where
        // each cell is marked when one
        // feature falls within it. This is not exactly the same than a pure
        // "min-distance" but is pretty close
        // and for large numbers of features is much faster than brute force search
        // of kd-trees.
        // (An intermediate approach would be the creation of a mask image updated
        // for each accepted feature, etc.)

        const bool do_filter_min_dist = options.FASTOptions.min_distance > 1;

        // Used half the min-distance since we'll later mark as occupied the ranges
        // [i-1,i+1] for a feature at "i"
        const unsigned int occupied_grid_cell_size =
                options.FASTOptions.min_distance / 2.0;
        const float occupied_grid_cell_size_inv = 1.0f / occupied_grid_cell_size;

        unsigned int grid_lx =
                !do_filter_min_dist
                        ? 1
                        : (unsigned int)(1 + inImg.getWidth() * occupied_grid_cell_size_inv);
        unsigned int grid_ly =
                !do_filter_min_dist
                        ? 1
                        : (unsigned int)(1 + inImg.getHeight() * occupied_grid_cell_size_inv);

        mrpt::math::CMatrixBool occupied_sections(
                grid_lx, grid_ly);  // See the comments above for an explanation.
        occupied_sections.fillAll(false);

        unsigned int nMax =
                (nDesiredFeatures != 0 && N > nDesiredFeatures) ? nDesiredFeatures : N;
        const int offset = (int)this->options.patchSize / 2 + 1;
        const size_t size_2 = options.patchSize / 2;
        const size_t imgH = inImg.getHeight();
        const size_t imgW = inImg.getWidth();
        unsigned int i = 0;
        unsigned int cont = 0;
        TFeatureID nextID = init_ID;

        if (!options.addNewFeatures) feats.clear();

        while (cont != nMax && i != N)
        {
                // Take the next feature fromt the ordered list of good features:
                const KeyPoint& kp = cv_feats[sorted_indices[i]];
                i++;

                // Patch out of the image??
                const int xBorderInf = (int)floor(kp.pt.x - size_2);
                const int xBorderSup = (int)floor(kp.pt.x + size_2);
                const int yBorderInf = (int)floor(kp.pt.y - size_2);
                const int yBorderSup = (int)floor(kp.pt.y + size_2);

                if (!(xBorderSup < (int)imgW && xBorderInf > 0 &&
                          yBorderSup < (int)imgH && yBorderInf > 0))
                        continue;  // nope, skip.

                if (do_filter_min_dist)
                {
                        // Check the min-distance:
                        const size_t section_idx_x =
                                size_t(kp.pt.x * occupied_grid_cell_size_inv);
                        const size_t section_idx_y =
                                size_t(kp.pt.y * occupied_grid_cell_size_inv);

                        if (occupied_sections(section_idx_x, section_idx_y))
                                continue;  // Already occupied! skip.

                        // Mark section as occupied
                        occupied_sections.set_unsafe(section_idx_x, section_idx_y, true);
                        if (section_idx_x > 0)
                                occupied_sections.set_unsafe(
                                        section_idx_x - 1, section_idx_y, true);
                        if (section_idx_y > 0)
                                occupied_sections.set_unsafe(
                                        section_idx_x, section_idx_y - 1, true);
                        if (section_idx_x < grid_lx - 1)
                                occupied_sections.set_unsafe(
                                        section_idx_x + 1, section_idx_y, true);
                        if (section_idx_y < grid_ly - 1)
                                occupied_sections.set_unsafe(
                                        section_idx_x, section_idx_y + 1, true);
                }

                // All tests passed: add new feature:
                CFeature::Ptr ft = mrpt::make_aligned_shared<CFeature>();
                ft->type = featFAST;
                ft->ID = nextID++;
                ft->x = kp.pt.x;
                ft->y = kp.pt.y;
                ft->response = kp.response;
                ft->orientation = kp.angle;
                ft->scale = kp.octave;
                ft->patchSize = options.patchSize;  // The size of the feature patch

                if (options.patchSize > 0)
                {
                        inImg.extract_patch(
                                ft->patch, round(ft->x) - offset, round(ft->y) - offset,
                                options.patchSize,
                                options.patchSize);  // Image patch surronding the feature
                }
                feats.push_back(ft);
                ++cont;
        }
// feats.resize( cont );  // JL: really needed???

#endif
#endif
        MRPT_END
}