Merge pull request #6933 from hrnr:gsoc_all

[GSOC] New camera model for stitching pipeline * implement estimateAffine2D estimates affine transformation using robust RANSAC method. * uses RANSAC framework in calib3d * includes accuracy test * uses SVD decomposition for solving 3 point equation * implement estimateAffinePartial2D estimates limited affine transformation * includes accuracy test * stitching: add affine matcher initial version of matcher that estimates affine transformation * stitching: added affine transform estimator initial version of estimator that simply chain transformations in homogeneous coordinates * calib3d: rename estimateAffine3D test test Calib3d_EstimateAffineTransform rename to Calib3d_EstimateAffine3D. This is more descriptive and prevents confusion with estimateAffine2D tests. * added perf test for estimateAffine functions tests both estimateAffine2D and estimateAffinePartial2D * calib3d: compare error in square in estimateAffine2D * incorporates fix from #6768 * rerun affine estimation on inliers * stitching: new API for parallel feature finding due to ABI breakage new functionality is added to `FeaturesFinder2`, `SurfFeaturesFinder2` and `OrbFeaturesFinder2` * stitching: add tests for parallel feature find API * perf test (about linear speed up) * accuracy test compares results with serial version * stitching: use dynamic_cast to overcome ABI issues adding parallel API to FeaturesFinder breaks ABI. This commit uses dynamic_cast and hardcodes thread-safe finders to avoid breaking ABI. This should be replaced by proper method similar to FeaturesMatcher on next ABI break. * use estimateAffinePartial2D in AffineBestOf2NearestMatcher * add constructor to AffineBestOf2NearestMatcher * allows to choose between full affine transform and partial affine transform. Other params are the as for BestOf2NearestMatcher * added protected field * samples: stitching_detailed support affine estimator and matcher * added new flags to choose matcher and estimator * stitching: rework affine matcher represent transformation in homogeneous coordinates affine matcher: remove duplicite code rework flow to get rid of duplicite code affine matcher: do not center points to (0, 0) it is not needed for affine model. it should not affect estimation in any way. affine matcher: remove unneeded cv namespacing * stitching: add stub bundle adjuster * adds stub bundle adjuster that does nothing * can be used in place of standard bundle adjusters to omit bundle adjusting step * samples: stitching detailed, support no budle adjust * uses new NoBundleAdjuster * added affine warper * uses R to get whole affine transformation and propagates rotation and translation to plane warper * add affine warper factory class * affine warper: compensate transformation * samples: stitching_detailed add support for affine warper * add Stitcher::create method this method follows similar constructor methods and returns smart pointer. This allows constructing Stitcher according to OpenCV guidelines. * supports multiple stitcher configurations (PANORAMA and SCANS) for convenient setup * returns cv::Ptr * stitcher: dynamicaly determine correct estimator we need to use affine estimator for affine matcher * preserves ABI (but add hints for ABI 4) * uses dynamic_cast hack to inject correct estimator * sample stitching: add support for multiple modes shows how to use different configurations of stitcher easily (panorama stitching and scans affine model) * stitcher: find features in parallel use new FeatureFinder API to find features in parallel. Parallelized using TBB. * stitching: disable parallel feature finding for OCL it does not bring much speedup to run features finder in parallel when OpenCL is enabled, because finder needs to wait for OCL device. Also, currently ORB is not thread-safe when OCL is enabled. * stitching: move matcher tests move matchers tests perf_stich.cpp -> perf_matchers.cpp * stitching: add affine stiching integration test test basic affine stitching (SCANS mode of stitcher) with images that have only translation between them * enable surf for stitching tests stitching.b12 test was failing with surf investigated the issue, surf is producing good result. Transformation is only slightly different from ORB, so that resulting pano does not exactly match ORB's result. That caused sanity check to fail. * added size checks similar to other tests * sanity check will be applied only for ORB * stitching: fix wrong estimator choice if case was exactly wrong, estimators were chosen wrong added logging for estimated transformation * enable surf for matchers stitching tests * enable SURF * rework sanity checking. Check estimated transform instead of matches. Est. transform should be more stable and comparable between SURF and ORB. * remove regression checking for VectorFeatures tests. It has a lot if data andtest is the same as previous except it test different vector size for performance, so sanity checking does not add any value here. Added basic sanity asserts instead. * stitching tests: allow relative error for transform * allows .01 relative error for estimated homography sanity check in stitching matchers tests * fix VS warning stitching tests: increase relative error increase relative error to make it pass on all platforms (results are still good). stitching test: allow bigger relative error transformation can differ in small values (with small absolute difference, but large relative difference). transformation output still looks usable for all platforms. This difference affects only mac and windows, linux passes fine with small difference. * stitching: add tests for affine matcher uses s1, s2 images. added also new sanity data. * stitching tests: use different data for matchers tests this data should yeild more stable transformation (it has much more matches, especially for surf). Sanity data regenerated. * stitching test: rework tests for matchers * separated rotation and translations as they are different by scale. * use appropriate absolute error for them separately. (relative error does not work for values near zero.) * stitching: fix affine warper compensation calculation of rotation and translation extracted for plane warper was wrong * stitching test: enable surf for opencl integration tests * enable SURF with correct guard (HAVE_OPENCV_XFEATURES2D) * add OPENCL guard and correct namespace as usual for opencl tests * stitching: add ocl accuracy test for affine warper test consistent results with ocl on and off * stitching: add affine warper ocl perf test add affine warper to existing warper perf tests. Added new sanity data. * stitching: do not overwrite inliers in affine matcher * estimation is run second time on inliers only, inliers produces in second run will not be therefore correct for all matches * calib3d: add Levenberg–Marquardt refining to estimateAffine2D* functions this adds affine Levenberg–Marquardt refining to estimateAffine2D functions similar to what is done in findHomography. implements Levenberg–Marquardt refinig for both full affine and partial affine transformations. * stitching: remove reestimation step in affine matcher reestimation step is not needed. estimateAffine2D* functions are running their own reestimation on inliers using the Levenberg-Marquardt algorithm, which is better than simply rerunning RANSAC on inliers. * implement partial affine bundle adjuster bundle adjuster that expect affine transform with 4DOF. Refines parameters for all cameras together. stitching: fix bug in BundleAdjusterAffinePartial * use the invers properly * use static buffer for invers to speed it up * samples: add affine bundle adjuster option to stitching_detailed * add support for using affine bundle adjuster with 4DOF * improve logging of initial intristics * sttiching: add affine bundle adjuster test * fix build warnings * stitching: increase limit on sanity check prevents spurious test failures on mac. values are still pretty fine. * stitching: set affine bundle adjuster for SCANS mode * fix bug with AffineBestOf2NearestMatcher (we want to select affine partial mode) * select right bundle adjuster * stitching: increase error bound for matcher tests * this prevents failure on mac. tranformation is still ok. * stitching: implement affine bundle adjuster * implements affine bundle adjuster that is using full affine transform * existing test case modified to test both affinePartial an full affine bundle adjuster * add stitching tutorial * show basic usage of stitching api (Stitcher class) * stitching: add more integration test for affine stitching * added new datasets to existing testcase * removed unused include * calib3d: move `haveCollinearPoints` to common header * added comment to make that this also checks too close points * calib3d: redone checkSubset for estimateAffine* callback * use common function to check collinearity * this also ensures that point will not be too close to each other * calib3d: change estimateAffine* functions API * more similar to `findHomography`, `findFundamentalMat`, `findEssentialMat` and similar * follows standard recommended semantic INPUTS, OUTPUTS, FLAGS * allows to disable refining * supported LMEDS robust method (tests yet to come) along with RANSAC * extended docs with some tips * calib3d: rewrite estimateAffine2D test * rewrite in googletest style * parametrize to test both robust methods (RANSAC and LMEDS) * get rid of boilerplate * calib3d: rework estimateAffinePartial2D test * rework in googletest style * add testing for LMEDS * calib3d: rework estimateAffine*2D perf test * test for LMEDS speed * test with/without Levenberg-Marquart * remove sanity checking (this is covered by accuracy tests) * calib3d: improve estimateAffine*2D tests * test transformations in loop * improves test by testing more potential transformations * calib3d: rewrite kernels for estimateAffine*2D functions * use analytical solution instead of SVD * this version is faster especially for smaller amount of points * calib3d: tune up perf of estimateAffine*2D functions * avoid copying inliers * avoid converting input points if not necessary * check only `from` point for collinearity, as `to` does not affect stability of transform * tutorials: add commands examples to stitching tutorials * add some examples how to run stitcher sample code * mention stitching_detailed.cpp * calib3d: change computeError for estimateAffine*2D * do error computing in floats instead of doubles this have required precision + we were storing the result in float anyway. This make code faster and allows auto-vectorization by smart compilers. * documentation: mention estimateAffine*2D function * refer to new functions on appropriate places * prefer estimateAffine*2D over estimateRigidTransform * stitching: add camera models documentations * mention camera models in module documentation to give user a better overview and reduce confusion
2026-07-30 15:53:03 +04:00 · 2016-10-22 18:10:42 +02:00
parent e77bc24b96
commit c17afe0fab
34 changed files with 2486 additions and 173 deletions
@@ -69,7 +69,29 @@ one can combine and use them separately.

 The implemented stitching pipeline is very similar to the one proposed in @cite BL07 .

-![image](StitchingPipeline.jpg)
+![stitching pipeline](StitchingPipeline.jpg)
+
+Camera models
+-------------
+
+There are currently 2 camera models implemented in stitching pipeline.
+
+- _Homography model_ expecting perspective transformations between images
+  implemented in @ref cv::detail::BestOf2NearestMatcher cv::detail::HomographyBasedEstimator
+  cv::detail::BundleAdjusterReproj cv::detail::BundleAdjusterRay
+- _Affine model_ expecting affine transformation with 6 DOF or 4 DOF implemented in
+  @ref cv::detail::AffineBestOf2NearestMatcher cv::detail::AffineBasedEstimator
+  cv::detail::BundleAdjusterAffine cv::detail::BundleAdjusterAffinePartial cv::AffineWarper
+
+Homography model is useful for creating photo panoramas captured by camera,
+while affine-based model can be used to stitch scans and object captured by
+specialized devices. Use @ref cv::Stitcher::create to get preconfigured pipeline for one
+of those models.
+
+@note
+Certain detailed settings of @ref cv::Stitcher might not make sense. Especially
+you should not mix classes implementing affine model and classes implementing
+Homography model, as they work with different transformations.

@{
    @defgroup stitching_match Features Finding and Images Matching
@@ -110,6 +132,22 @@ public:
        ERR_HOMOGRAPHY_EST_FAIL = 2,
        ERR_CAMERA_PARAMS_ADJUST_FAIL = 3
    };
+    enum Mode
+    {
+        /** Mode for creating photo panoramas. Expects images under perspective
+        transformation and projects resulting pano to sphere.
+
+        @sa detail::BestOf2NearestMatcher SphericalWarper
+        */
+        PANORAMA = 0,
+        /** Mode for composing scans. Expects images under affine transformation does
+        not compensate exposure by default.
+
+        @sa detail::AffineBestOf2NearestMatcher AffineWarper
+        */
+        SCANS = 1,
+
+    };

   // Stitcher() {}
    /** @brief Creates a stitcher with the default parameters.
@@ -118,6 +156,15 @@ public:
    @return Stitcher class instance.
     */
    static Stitcher createDefault(bool try_use_gpu = false);
+    /** @brief Creates a Stitcher configured in one of the stitching modes.
+
+    @param mode Scenario for stitcher operation. This is usually determined by source of images
+    to stitch and their transformation. Default parameters will be chosen for operation in given
+    scenario.
+    @param try_use_gpu Flag indicating whether GPU should be used whenever it's possible.
+    @return Stitcher class instance.
+     */
+    static Ptr<Stitcher> create(Mode mode = PANORAMA, bool try_use_gpu = false);

    CV_WRAP double registrationResol() const { return registr_resol_; }
    CV_WRAP void setRegistrationResol(double resol_mpx) { registr_resol_ = resol_mpx; }
@@ -159,6 +206,13 @@ public:
    void setBundleAdjuster(Ptr<detail::BundleAdjusterBase> bundle_adjuster)
        { bundle_adjuster_ = bundle_adjuster; }

+    /* TODO OpenCV ABI 4.x
+    Ptr<detail::Estimator> estimator() { return estimator_; }
+    const Ptr<detail::Estimator> estimator() const { return estimator_; }
+    void setEstimator(Ptr<detail::Estimator> estimator)
+        { estimator_ = estimator; }
+    */
+
    Ptr<WarperCreator> warper() { return warper_; }
    const Ptr<WarperCreator> warper() const { return warper_; }
    void setWarper(Ptr<WarperCreator> creator) { warper_ = creator; }
@@ -233,6 +287,9 @@ private:
    Ptr<detail::FeaturesMatcher> features_matcher_;
    cv::UMat matching_mask_;
    Ptr<detail::BundleAdjusterBase> bundle_adjuster_;
+    /* TODO OpenCV ABI 4.x
+    Ptr<detail::Estimator> estimator_;
+    */
    bool do_wave_correct_;
    detail::WaveCorrectKind wave_correct_kind_;
    Ptr<WarperCreator> warper_;
@@ -83,9 +83,27 @@ public:
    @sa detail::ImageFeatures, Rect_
    */
    void operator ()(InputArray image, ImageFeatures &features, const std::vector<cv::Rect> &rois);
+    /** @brief Finds features in the given images in parallel.
+
+    @param images Source images
+    @param features Found features for each image
+    @param rois Regions of interest for each image
+
+    @sa detail::ImageFeatures, Rect_
+    */
+    void operator ()(InputArrayOfArrays images, std::vector<ImageFeatures> &features,
+                     const std::vector<std::vector<cv::Rect> > &rois);
+    /** @overload */
+    void operator ()(InputArrayOfArrays images, std::vector<ImageFeatures> &features);
    /** @brief Frees unused memory allocated before if there is any. */
    virtual void collectGarbage() {}

+    /* TODO OpenCV ABI 4.x
+    reimplement this as public method similar to FeaturesMatcher and remove private function hack
+    @return True, if it's possible to use the same finder instance in parallel, false otherwise
+    bool isThreadSafe() const { return is_thread_safe_; }
+    */
+
 protected:
    /** @brief This method must implement features finding logic in order to make the wrappers
    detail::FeaturesFinder::operator()_ work.
@@ -95,6 +113,10 @@ protected:

    @sa detail::ImageFeatures */
    virtual void find(InputArray image, ImageFeatures &features) = 0;
+    /** @brief uses dynamic_cast to determine thread-safety
+    @return True, if it's possible to use the same finder instance in parallel, false otherwise
+    */
+    bool isThreadSafe() const;
 };

 /** @brief SURF features finder.
@@ -152,7 +174,6 @@ private:
    Ptr<AKAZE> akaze;
 };

-
 #ifdef HAVE_OPENCV_XFEATURES2D
 class CV_EXPORTS SurfFeaturesFinderGpu : public FeaturesFinder
 {
@@ -177,7 +198,10 @@ private:

 /** @brief Structure containing information about matches between two images.

-It's assumed that there is a homography between those images.
+It's assumed that there is a transformation between those images. Transformation may be
+homography or affine transformation based on selected matcher.
+
+@sa detail::FeaturesMatcher
 */
 struct CV_EXPORTS MatchesInfo
 {
@@ -189,7 +213,7 @@ struct CV_EXPORTS MatchesInfo
    std::vector<DMatch> matches;
    std::vector<uchar> inliers_mask;    //!< Geometrically consistent matches mask
    int num_inliers;                    //!< Number of geometrically consistent matches
-    Mat H;                              //!< Estimated homography
+    Mat H;                              //!< Estimated transformation
    double confidence;                  //!< Confidence two images are from the same panorama
 };

@@ -288,6 +312,41 @@ protected:
    int range_width_;
 };

+/** @brief Features matcher similar to cv::detail::BestOf2NearestMatcher which
+finds two best matches for each feature and leaves the best one only if the
+ratio between descriptor distances is greater than the threshold match_conf.
+
+Unlike cv::detail::BestOf2NearestMatcher this matcher uses affine
+transformation (affine trasformation estimate will be placed in matches_info).
+
+@sa cv::detail::FeaturesMatcher cv::detail::BestOf2NearestMatcher
+ */
+class CV_EXPORTS AffineBestOf2NearestMatcher : public BestOf2NearestMatcher
+{
+public:
+    /** @brief Constructs a "best of 2 nearest" matcher that expects affine trasformation
+    between images
+
+    @param full_affine whether to use full affine transformation with 6 degress of freedom or reduced
+    transformation with 4 degrees of freedom using only rotation, translation and uniform scaling
+    @param try_use_gpu Should try to use GPU or not
+    @param match_conf Match distances ration threshold
+    @param num_matches_thresh1 Minimum number of matches required for the 2D affine transform
+    estimation used in the inliers classification step
+
+    @sa cv::estimateAffine2D cv::estimateAffinePartial2D
+     */
+    AffineBestOf2NearestMatcher(bool full_affine = false, bool try_use_gpu = false,
+                                float match_conf = 0.3f, int num_matches_thresh1 = 6) :
+        BestOf2NearestMatcher(try_use_gpu, match_conf, num_matches_thresh1, num_matches_thresh1),
+        full_affine_(full_affine) {}
+
+protected:
+    void match(const ImageFeatures &features1, const ImageFeatures &features2, MatchesInfo &matches_info);
+
+    bool full_affine_;
+};
+
 //! @} stitching_match

 } // namespace detail
@@ -109,6 +109,21 @@ private:
    bool is_focals_estimated_;
 };

+/** @brief Affine transformation based estimator.
+
+This estimator uses pairwise tranformations estimated by matcher to estimate
+final transformation for each camera.
+
+@sa cv::detail::HomographyBasedEstimator
+ */
+class CV_EXPORTS AffineBasedEstimator : public Estimator
+{
+private:
+    virtual bool estimate(const std::vector<ImageFeatures> &features,
+                          const std::vector<MatchesInfo> &pairwise_matches,
+                          std::vector<CameraParams> &cameras);
+};
+
 /** @brief Base class for all camera parameters refinement methods.
 */
 class CV_EXPORTS BundleAdjusterBase : public Estimator
@@ -195,6 +210,26 @@ protected:
 };


+/** @brief Stub bundle adjuster that does nothing.
+ */
+class CV_EXPORTS NoBundleAdjuster : public BundleAdjusterBase
+{
+public:
+    NoBundleAdjuster() : BundleAdjusterBase(0, 0) {}
+
+private:
+    bool estimate(const std::vector<ImageFeatures> &, const std::vector<MatchesInfo> &,
+                  std::vector<CameraParams> &)
+    {
+        return true;
+    }
+    void setUpInitialCameraParams(const std::vector<CameraParams> &) {}
+    void obtainRefinedCameraParams(std::vector<CameraParams> &) const {}
+    void calcError(Mat &) {}
+    void calcJacobian(Mat &) {}
+};
+
+
 /** @brief Implementation of the camera parameters refinement algorithm which minimizes sum of the reprojection
 error squares

@@ -236,6 +271,54 @@ private:
 };


+/** @brief Bundle adjuster that expects affine transformation
+represented in homogeneous coordinates in R for each camera param. Implements
+camera parameters refinement algorithm which minimizes sum of the reprojection
+error squares
+
+It estimates all transformation parameters. Refinement mask is ignored.
+
+@sa AffineBasedEstimator AffineBestOf2NearestMatcher BundleAdjusterAffinePartial
+ */
+class CV_EXPORTS BundleAdjusterAffine : public BundleAdjusterBase
+{
+public:
+    BundleAdjusterAffine() : BundleAdjusterBase(6, 2) {}
+
+private:
+    void setUpInitialCameraParams(const std::vector<CameraParams> &cameras);
+    void obtainRefinedCameraParams(std::vector<CameraParams> &cameras) const;
+    void calcError(Mat &err);
+    void calcJacobian(Mat &jac);
+
+    Mat err1_, err2_;
+};
+
+
+/** @brief Bundle adjuster that expects affine transformation with 4 DOF
+represented in homogeneous coordinates in R for each camera param. Implements
+camera parameters refinement algorithm which minimizes sum of the reprojection
+error squares
+
+It estimates all transformation parameters. Refinement mask is ignored.
+
+@sa AffineBasedEstimator AffineBestOf2NearestMatcher BundleAdjusterAffine
+ */
+class CV_EXPORTS BundleAdjusterAffinePartial : public BundleAdjusterBase
+{
+public:
+    BundleAdjusterAffinePartial() : BundleAdjusterBase(4, 2) {}
+
+private:
+    void setUpInitialCameraParams(const std::vector<CameraParams> &cameras);
+    void obtainRefinedCameraParams(std::vector<CameraParams> &cameras) const;
+    void calcError(Mat &err);
+    void calcJacobian(Mat &jac);
+
+    Mat err1_, err2_;
+};
+
+
 enum WaveCorrectKind
 {
    WAVE_CORRECT_HORIZ,
@@ -205,6 +205,34 @@ protected:
 };


+/** @brief Affine warper that uses rotations and translations
+
+ Uses affine transformation in homogeneous coordinates to represent both rotation and
+ translation in camera rotation matrix.
+ */
+class CV_EXPORTS AffineWarper : public PlaneWarper
+{
+public:
+    /** @brief Construct an instance of the affine warper class.
+
+    @param scale Projected image scale multiplier
+     */
+    AffineWarper(float scale = 1.f) : PlaneWarper(scale) {}
+
+    Point2f warpPoint(const Point2f &pt, InputArray K, InputArray R);
+    Rect buildMaps(Size src_size, InputArray K, InputArray R, OutputArray xmap, OutputArray ymap);
+    Point warp(InputArray src, InputArray K, InputArray R,
+               int interp_mode, int border_mode, OutputArray dst);
+    Rect warpRoi(Size src_size, InputArray K, InputArray R);
+
+protected:
+    /** @brief Extracts rotation and translation matrices from matrix H representing
+        affine transformation in homogeneous coordinates
+     */
+    void getRTfromHomogeneous(InputArray H, Mat &R, Mat &T);
+};
+
+
 struct CV_EXPORTS SphericalProjector : ProjectorBase
 {
    void mapForward(float x, float y, float &u, float &v);
@@ -68,6 +68,15 @@ public:
    Ptr<detail::RotationWarper> create(float scale) const { return makePtr<detail::PlaneWarper>(scale); }
 };

+/** @brief Affine warper factory class.
+  @sa detail::AffineWarper
+ */
+class AffineWarper : public WarperCreator
+{
+public:
+    Ptr<detail::RotationWarper> create(float scale) const { return makePtr<detail::AffineWarper>(scale); }
+};
+
 /** @brief Cylindrical warper factory class.
@sa detail::CylindricalWarper
 */