Merge remote-tracking branch 'upstream/3.4' into merge-3.4

apps/traincascade/old_ml.hpp

@@ -38,8 +38,8 @@
 //
 //M*/
 
-#ifndef OPENCV_ML_HPP
+#ifndef OPENCV_OLD_ML_HPP
-#define OPENCV_ML_HPP
+#define OPENCV_OLD_ML_HPP
 
 #ifdef __cplusplus
 #  include "opencv2/core.hpp"
@@ -2038,6 +2038,6 @@ template<> struct DefaultDeleter<CvDTreeSplit>{ void operator ()(CvDTreeSplit* o
 }
 
 #endif // __cplusplus
-#endif // OPENCV_ML_HPP
+#endif // OPENCV_OLD_ML_HPP
 
 /* End of file. */

cmake/OpenCVDetectCUDA.cmake

@@ -87,6 +87,8 @@ if(CUDA_FOUND)
     set(__cuda_arch_bin "6.0 6.1")
   elseif(CUDA_GENERATION STREQUAL "Volta")
     set(__cuda_arch_bin "7.0")
+  elseif(CUDA_GENERATION STREQUAL "Turing")
+    set(__cuda_arch_bin "7.5")
   elseif(CUDA_GENERATION STREQUAL "Auto")
     execute_process( COMMAND ${DETECT_ARCHS_COMMAND}
                      WORKING_DIRECTORY "${CMAKE_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/CMakeTmp/"
@@ -113,7 +115,7 @@ if(CUDA_FOUND)
       string(REGEX REPLACE ".*\n" "" _nvcc_out "${_nvcc_out}") #Strip leading warning messages, if any
       if(NOT _nvcc_res EQUAL 0)
         message(STATUS "Automatic detection of CUDA generation failed. Going to build for all known architectures.")
-        set(__cuda_arch_bin "5.3 6.2 7.0")
+        set(__cuda_arch_bin "5.3 6.2 7.0 7.5")
       else()
         set(__cuda_arch_bin "${_nvcc_out}")
         string(REPLACE "2.1" "2.1(2.0)" __cuda_arch_bin "${__cuda_arch_bin}")
@@ -122,8 +124,10 @@ if(CUDA_FOUND)
     else()
       if(${CUDA_VERSION} VERSION_LESS "9.0")
         set(__cuda_arch_bin "2.0 3.0 3.5 3.7 5.0 5.2 6.0 6.1")
-      else()
+      elseif(${CUDA_VERSION} VERSION_LESS "10.0")
         set(__cuda_arch_bin "3.0 3.5 3.7 5.0 5.2 6.0 6.1 7.0")
+      else()
+        set(__cuda_arch_bin "3.0 3.5 3.7 5.0 5.2 6.0 6.1 7.0 7.5")
       endif()
     endif()
   endif()

modules/core/include/opencv2/core/types_c.h

@@ -1353,7 +1353,7 @@ CvSlice;
 
 CV_INLINE  CvSlice  cvSlice( int start, int end )
 {
-#if !(defined(CV__ENABLE_C_API_CTORS) && defined(__cplusplus))
+#if !(defined(CV__ENABLE_C_API_CTORS) && defined(__cplusplus) && !defined(__CUDACC__))
     CvSlice slice = { start, end };
 #else
     CvSlice slice(start, end);

modules/core/src/copy.cpp

@@ -729,7 +729,7 @@ static bool ipp_flip(Mat &src, Mat &dst, int flip_mode)
 
         CV_INSTRUMENT_FUN_IPP(::ipp::iwiMirror, iwSrc, iwDst, ippMode);
     }
-    catch(::ipp::IwException)
+    catch(const ::ipp::IwException &)
     {
         return false;
     }

modules/core/test/test_ds.cpp

@@ -1032,7 +1032,7 @@ void Core_SeqBaseTest::run( int )
                 cvClearMemStorage( storage );
         }
     }
-    catch(int)
+    catch(const int &)
     {
     }
 }
@@ -1200,7 +1200,7 @@ void Core_SeqSortInvTest::run( int )
             storage.release();
         }
     }
-    catch (int)
+    catch (const int &)
     {
     }
 }
@@ -1416,7 +1416,7 @@ void Core_SetTest::run( int )
             storage.release();
         }
     }
-    catch(int)
+    catch(const int &)
     {
     }
 }
@@ -1859,7 +1859,7 @@ void Core_GraphTest::run( int )
             storage.release();
         }
     }
-    catch(int)
+    catch(const int &)
     {
     }
 }
@@ -2121,7 +2121,7 @@ void Core_GraphScanTest::run( int )
             storage.release();
         }
     }
-    catch(int)
+    catch(const int &)
     {
     }
 }

modules/dnn/src/op_inf_engine.cpp

@@ -331,8 +331,18 @@ InferenceEngine::StatusCode InfEngineBackendNet::setBatchSize(size_t size, Infer
 
 size_t InfEngineBackendNet::getBatchSize() const CV_NOEXCEPT
 {
-    CV_Error(Error::StsNotImplemented, "");
+    size_t batchSize = 0;
-    return 0;
+    for (const auto& inp : inputs)
+    {
+        CV_Assert(inp.second);
+        std::vector<size_t> dims = inp.second->getDims();
+        CV_Assert(!dims.empty());
+        if (batchSize != 0)
+            CV_Assert(batchSize == dims.back());
+        else
+            batchSize = dims.back();
+    }
+    return batchSize;
 }
 
 #if INF_ENGINE_VER_MAJOR_GT(INF_ENGINE_RELEASE_2018R2)

modules/dnn/test/test_torch_importer.cpp

@@ -287,6 +287,46 @@ TEST_P(Test_Torch_nets, OpenFace_accuracy)
     normAssert(out, outRef, "", default_l1, default_lInf);
 }
 
+static Mat getSegmMask(const Mat& scores)
+{
+    const int rows = scores.size[2];
+    const int cols = scores.size[3];
+    const int numClasses = scores.size[1];
+
+    Mat maxCl = Mat::zeros(rows, cols, CV_8UC1);
+    Mat maxVal(rows, cols, CV_32FC1, Scalar(0));
+    for (int ch = 0; ch < numClasses; ch++)
+    {
+        for (int row = 0; row < rows; row++)
+        {
+            const float *ptrScore = scores.ptr<float>(0, ch, row);
+            uint8_t *ptrMaxCl = maxCl.ptr<uint8_t>(row);
+            float *ptrMaxVal = maxVal.ptr<float>(row);
+            for (int col = 0; col < cols; col++)
+            {
+                if (ptrScore[col] > ptrMaxVal[col])
+                {
+                    ptrMaxVal[col] = ptrScore[col];
+                    ptrMaxCl[col] = (uchar)ch;
+                }
+            }
+        }
+    }
+    return maxCl;
+}
+
+// Computer per-class intersection over union metric.
+static void normAssertSegmentation(const Mat& ref, const Mat& test)
+{
+    CV_Assert_N(ref.dims == 4, test.dims == 4);
+    const int numClasses = ref.size[1];
+    CV_Assert(numClasses == test.size[1]);
+
+    Mat refMask = getSegmMask(ref);
+    Mat testMask = getSegmMask(test);
+    EXPECT_EQ(countNonZero(refMask != testMask), 0);
+}
+
 TEST_P(Test_Torch_nets, ENet_accuracy)
 {
     checkBackend();
@@ -313,14 +353,16 @@ TEST_P(Test_Torch_nets, ENet_accuracy)
     // Due to numerical instability in Pooling-Unpooling layers (indexes jittering)
     // thresholds for ENet must be changed. Accuracy of results was checked on
     // Cityscapes dataset and difference in mIOU with Torch is 10E-4%
-    normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.5);
+    normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.552);
+    normAssertSegmentation(ref, out);
 
     const int N = 3;
     for (int i = 0; i < N; i++)
     {
         net.setInput(inputBlob, "");
         Mat out = net.forward();
-        normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.5);
+        normAssert(ref, out, "", 0.00044, /*target == DNN_TARGET_CPU ? 0.453 : */0.552);
+        normAssertSegmentation(ref, out);
     }
 }
 

modules/features2d/src/kaze/KAZEConfig.h

@@ -5,8 +5,8 @@
  * @author Pablo F. Alcantarilla
  */
 
-#ifndef __OPENCV_FEATURES_2D_AKAZE_CONFIG_H__
+#ifndef __OPENCV_FEATURES_2D_KAZE_CONFIG_H__
-#define __OPENCV_FEATURES_2D_AKAZE_CONFIG_H__
+#define __OPENCV_FEATURES_2D_KAZE_CONFIG_H__
 
 // OpenCV Includes
 #include "../precomp.hpp"

modules/flann/include/opencv2/flann.hpp

@@ -103,6 +103,58 @@ using ::cvflann::KL_Divergence;
 
 /** @brief The FLANN nearest neighbor index class. This class is templated with the type of elements for which
 the index is built.
+
+`Distance` functor specifies the metric to be used to calculate the distance between two points.
+There are several `Distance` functors that are readily available:
+
+@link cvflann::L2_Simple cv::flann::L2_Simple @endlink- Squared Euclidean distance functor.
+This is the simpler, unrolled version. This is preferable for very low dimensionality data (eg 3D points)
+
+@link cvflann::L2 cv::flann::L2 @endlink- Squared Euclidean distance functor, optimized version.
+
+@link cvflann::L1 cv::flann::L1 @endlink - Manhattan distance functor, optimized version.
+
+@link cvflann::MinkowskiDistance cv::flann::MinkowskiDistance @endlink -  The Minkowsky distance functor.
+This is highly optimised with loop unrolling.
+The computation of squared root at the end is omitted for efficiency.
+
+@link cvflann::MaxDistance cv::flann::MaxDistance @endlink - The max distance functor. It computes the
+maximum distance between two vectors. This distance is not a valid kdtree distance, it's not
+dimensionwise additive.
+
+@link cvflann::HammingLUT cv::flann::HammingLUT @endlink -  %Hamming distance functor. It counts the bit
+differences between two strings using a lookup table implementation.
+
+@link cvflann::Hamming cv::flann::Hamming @endlink - %Hamming distance functor. Population count is
+performed using library calls, if available. Lookup table implementation is used as a fallback.
+
+@link cvflann::Hamming2 cv::flann::Hamming2 @endlink- %Hamming distance functor. Population count is
+implemented in 12 arithmetic operations (one of which is multiplication).
+
+@link cvflann::HistIntersectionDistance cv::flann::HistIntersectionDistance @endlink - The histogram
+intersection distance functor.
+
+@link cvflann::HellingerDistance cv::flann::HellingerDistance @endlink - The Hellinger distance functor.
+
+@link cvflann::ChiSquareDistance cv::flann::ChiSquareDistance @endlink - The chi-square distance functor.
+
+@link cvflann::KL_Divergence cv::flann::KL_Divergence @endlink - The Kullback-Leibler divergence functor.
+
+Although the provided implementations cover a vast range of cases, it is also possible to use
+a custom implementation. The distance functor is a class whose `operator()` computes the distance
+between two features. If the distance is also a kd-tree compatible distance, it should also provide an
+`accum_dist()` method that computes the distance between individual feature dimensions.
+
+In addition to `operator()` and `accum_dist()`, a distance functor should also define the
+`ElementType` and the `ResultType` as the types of the elements it operates on and the type of the
+result it computes. If a distance functor can be used as a kd-tree distance (meaning that the full
+distance between a pair of features can be accumulated from the partial distances between the
+individual dimensions) a typedef `is_kdtree_distance` should be present inside the distance functor.
+If the distance is not a kd-tree distance, but it's a distance in a vector space (the individual
+dimensions of the elements it operates on can be accessed independently) a typedef
+`is_vector_space_distance` should be defined inside the functor. If neither typedef is defined, the
+distance is assumed to be a metric distance and will only be used with indexes operating on
+generic metric distances.
  */
 template <typename Distance>
 class GenericIndex
@@ -217,6 +269,17 @@ public:
                        std::vector<DistanceType>& dists, int knn, const ::cvflann::SearchParams& params);
         void knnSearch(const Mat& queries, Mat& indices, Mat& dists, int knn, const ::cvflann::SearchParams& params);
 
+        /** @brief Performs a radius nearest neighbor search for a given query point using the index.
+
+        @param query The query point.
+        @param indices Vector that will contain the indices of the nearest neighbors found.
+        @param dists Vector that will contain the distances to the nearest neighbors found. It has the same
+        number of elements as indices.
+        @param radius The search radius.
+        @param params SearchParams
+
+        This function returns the number of nearest neighbors found.
+        */
         int radiusSearch(const std::vector<ElementType>& query, std::vector<int>& indices,
                          std::vector<DistanceType>& dists, DistanceType radius, const ::cvflann::SearchParams& params);
         int radiusSearch(const Mat& query, Mat& indices, Mat& dists,

modules/imgproc/src/canny.cpp

@@ -93,7 +93,7 @@ static bool ipp_Canny(const Mat& src , const Mat& dx_, const Mat& dy_, Mat& dst,
 
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterCannyDeriv, iwSrcDx, iwSrcDy, iwDst, low, high, ::ipp::IwiFilterCannyDerivParams(norm));
         }
-        catch (::ipp::IwException ex)
+        catch (const ::ipp::IwException &)
         {
             return false;
         }
@@ -119,7 +119,7 @@ static bool ipp_Canny(const Mat& src , const Mat& dx_, const Mat& dy_, Mat& dst,
 
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterCanny, iwSrc, iwDst, low, high, ::ipp::IwiFilterCannyParams(ippFilterSobel, kernel, norm), ippBorderRepl);
         }
-        catch (::ipp::IwException)
+        catch (const ::ipp::IwException &)
         {
             return false;
         }

modules/imgproc/src/deriv.cpp

@@ -337,7 +337,7 @@ static bool ipp_Deriv(InputArray _src, OutputArray _dst, int dx, int dy, int ksi
         if(useScale)
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiScale, iwDstProc, iwDst, scale, delta, ::ipp::IwiScaleParams(ippAlgHintFast));
     }
-    catch (::ipp::IwException)
+    catch (const ::ipp::IwException &)
     {
         return false;
     }
@@ -765,7 +765,7 @@ static bool ipp_Laplacian(InputArray _src, OutputArray _dst, int ksize, double s
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiScale, iwDstProc, iwDst, scale, delta);
 
     }
-    catch (::ipp::IwException ex)
+    catch (const ::ipp::IwException &)
     {
         return false;
     }

modules/imgproc/src/morph.cpp

@@ -1299,7 +1299,7 @@ static bool ippMorph(int op, int src_type, int dst_type,
                 CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterMorphology, iwSrc, iwDst, morphType, iwMask, ::ipp::IwDefault(), iwBorderType);
         }
     }
-    catch(::ipp::IwException ex)
+    catch(const ::ipp::IwException &)
     {
         return false;
     }

modules/imgproc/src/resize.cpp

@@ -3241,7 +3241,7 @@ public:
             ::ipp::IwiTile tile = ::ipp::IwiRoi(0, range.start, m_dst.m_size.width, range.end - range.start);
             CV_INSTRUMENT_FUN_IPP(iwiResize, m_src, m_dst, ippBorderRepl, tile);
         }
-        catch(::ipp::IwException)
+        catch(const ::ipp::IwException &)
         {
             m_ok = false;
             return;
@@ -3291,7 +3291,7 @@ public:
             ::ipp::IwiTile tile = ::ipp::IwiRoi(0, range.start, m_dst.m_size.width, range.end - range.start);
             CV_INSTRUMENT_FUN_IPP(iwiWarpAffine, m_src, m_dst, tile);
         }
-        catch(::ipp::IwException)
+        catch(const ::ipp::IwException &)
         {
             m_ok = false;
             return;
@@ -3387,7 +3387,7 @@ static bool ipp_resize(const uchar * src_data, size_t src_step, int src_width, i
         if(!ok)
             return false;
     }
-    catch(::ipp::IwException)
+    catch(const ::ipp::IwException &)
     {
         return false;
     }

modules/imgproc/src/smooth.cpp

@@ -1510,7 +1510,7 @@ static bool ipp_boxfilter(Mat &src, Mat &dst, Size ksize, Point anchor, bool nor
 
         CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterBox, iwSrc, iwDst, iwKSize, ::ipp::IwDefault(), ippBorder);
     }
-    catch (::ipp::IwException)
+    catch (const ::ipp::IwException &)
     {
         return false;
     }
@@ -4000,7 +4000,7 @@ public:
             ::ipp::IwiTile tile = ::ipp::IwiRoi(0, range.start, m_dst.m_size.width, range.end - range.start);
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterGaussian, m_src, m_dst, m_kernelSize, m_sigma, ::ipp::IwDefault(), m_border, tile);
         }
-        catch(::ipp::IwException e)
+        catch(const ::ipp::IwException &)
         {
             *m_pOk = false;
             return;
@@ -4067,7 +4067,7 @@ static bool ipp_GaussianBlur(InputArray _src, OutputArray _dst, Size ksize,
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterGaussian, iwSrc, iwDst, ksize.width, sigma1, ::ipp::IwDefault(), ippBorder);
         }
     }
-    catch (::ipp::IwException ex)
+    catch (const ::ipp::IwException &)
     {
         return false;
     }
@@ -5873,7 +5873,7 @@ public:
             ::ipp::IwiTile tile = ::ipp::IwiRoi(0, range.start, dst.m_size.width, range.end - range.start);
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterBilateral, src, dst, radius, valSquareSigma, posSquareSigma, ::ipp::IwDefault(), borderType, tile);
         }
-        catch(::ipp::IwException)
+        catch(const ::ipp::IwException &)
         {
             *pOk = false;
             return;
@@ -5928,7 +5928,7 @@ static bool ipp_bilateralFilter(Mat &src, Mat &dst, int d, double sigmaColor, do
             CV_INSTRUMENT_FUN_IPP(::ipp::iwiFilterBilateral, iwSrc, iwDst, radius, valSquareSigma, posSquareSigma, ::ipp::IwDefault(), ippBorder);
         }
     }
-    catch (::ipp::IwException)
+    catch (const ::ipp::IwException &)
     {
         return false;
     }

modules/stitching/perf/perf_matchers.cpp

@@ -103,7 +103,7 @@ PERF_TEST_P( match, bestOf2Nearest, TEST_DETECTORS)
     Mat R (pairwise_matches.H, Range::all(), Range(0, 2));
     // separate transform matrix, use lower error on rotations
     SANITY_CHECK(dist, 1., ERROR_ABSOLUTE);
-    SANITY_CHECK(R, .015, ERROR_ABSOLUTE);
+    SANITY_CHECK(R, .06, ERROR_ABSOLUTE);
 }
 
 PERF_TEST_P( matchVector, bestOf2NearestVectorFeatures, testing::Combine(

modules/videoio/src/cap_ffmpeg_impl.hpp

@@ -1519,8 +1519,6 @@ static AVFrame * icv_alloc_picture_FFMPEG(int pix_fmt, int width, int height, bo
         _opencv_ffmpeg_av_image_fill_arrays(picture, picture_buf,
                        (AVPixelFormat) pix_fmt, width, height);
     }
-    else {
-    }
 
     return picture;
 }