diff --git a/CMakeLists.txt b/CMakeLists.txt
index f4dbfc4b69..a177356b43 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -769,8 +769,9 @@ if(HAVE_CUDA)
   status("")
   status("  NVIDIA CUDA")
 
-  status("    Use CUFFT:"            HAVE_CUFFT  THEN YES ELSE NO)
-  status("    Use CUBLAS:"           HAVE_CUBLAS THEN YES ELSE NO)
+  status("    Use CUFFT:"            HAVE_CUFFT   THEN YES ELSE NO)
+  status("    Use CUBLAS:"           HAVE_CUBLAS  THEN YES ELSE NO)
+  status("    USE NVCUVID:"          HAVE_NVCUVID THEN YES ELSE NO)
   status("    NVIDIA GPU arch:"      ${OPENCV_CUDA_ARCH_BIN})
   status("    NVIDIA PTX archs:"     ${OPENCV_CUDA_ARCH_PTX})
   status("    Use fast math:"        CUDA_FAST_MATH THEN YES ELSE NO)
diff --git a/modules/gpu/CMakeLists.txt b/modules/gpu/CMakeLists.txt
index 2f62826dd5..0c9f709b15 100644
--- a/modules/gpu/CMakeLists.txt
+++ b/modules/gpu/CMakeLists.txt
@@ -5,7 +5,7 @@ endif()
 set(the_description "GPU-accelerated Computer Vision")
 ocv_add_module(gpu opencv_imgproc opencv_calib3d opencv_objdetect opencv_video opencv_nonfree opencv_photo opencv_legacy)
 
-ocv_module_include_directories("${CMAKE_CURRENT_SOURCE_DIR}/src/cuda" "${CMAKE_CURRENT_SOURCE_DIR}/../highgui/src")
+ocv_module_include_directories("${CMAKE_CURRENT_SOURCE_DIR}/src/cuda")
 
 file(GLOB lib_hdrs               "include/opencv2/${name}/*.hpp"               "include/opencv2/${name}/*.h")
 file(GLOB lib_device_hdrs        "include/opencv2/${name}/device/*.hpp"        "include/opencv2/${name}/device/*.h")
diff --git a/modules/gpu/src/ffmpeg_video_source.cpp b/modules/gpu/src/ffmpeg_video_source.cpp
index 42327fef38..dbbe780d0a 100644
--- a/modules/gpu/src/ffmpeg_video_source.cpp
+++ b/modules/gpu/src/ffmpeg_video_source.cpp
@@ -45,9 +45,9 @@
 #if defined(HAVE_CUDA) && defined(HAVE_NVCUVID)
 
 #if defined(HAVE_FFMPEG) && defined(BUILD_SHARED_LIBS)
-    #include "cap_ffmpeg_impl.hpp"
+    #include "../src/cap_ffmpeg_impl.hpp"
 #else
-    #include "cap_ffmpeg_api.hpp"
+    #include "../src/cap_ffmpeg_api.hpp"
 #endif
 
 namespace
diff --git a/modules/gpu/src/gpu_init.cpp b/modules/gpu/src/gpu_init.cpp
index cffacb833e..8ed93651ad 100644
--- a/modules/gpu/src/gpu_init.cpp
+++ b/modules/gpu/src/gpu_init.cpp
@@ -40,7 +40,7 @@
 //
 //M*/
 
-#include <precomp.hpp>
+#include "precomp.hpp"
 
 namespace cv { namespace gpu
 {
diff --git a/modules/gpu/src/softcascade.cpp b/modules/gpu/src/softcascade.cpp
index 511c4525d7..561d4e44e1 100644
--- a/modules/gpu/src/softcascade.cpp
+++ b/modules/gpu/src/softcascade.cpp
@@ -40,7 +40,7 @@
 //
 //M*/
 
-#include <precomp.hpp>
+#include "precomp.hpp"
 
 #if !defined (HAVE_CUDA)
 cv::gpu::SCascade::SCascade(const double, const double, const int, const int) { throw_nogpu(); }
@@ -60,7 +60,7 @@ cv::Ptr<cv::gpu::ChannelsProcessor> cv::gpu::ChannelsProcessor::create(const int
 { throw_nogpu(); return cv::Ptr<cv::gpu::ChannelsProcessor>(0); }
 
 #else
-# include <icf.hpp>
+# include "icf.hpp"
 
 cv::gpu::device::icf::Level::Level(int idx, const Octave& oct, const float scale, const int w, const int h)
 :  octave(idx), step(oct.stages), relScale(scale / oct.scale)
@@ -95,14 +95,6 @@ namespace icf {
     void shrink(const cv::gpu::PtrStepSzb& channels, cv::gpu::PtrStepSzb shrunk);
 }
 
-// namespace imgproc {
-//     void shfl_integral_gpu_buffered(PtrStepSzb, PtrStepSz<uint4>, PtrStepSz<unsigned int>, int, cudaStream_t);
-
-//     template <typename T>
-//     void resize_gpu(PtrStepSzb src, PtrStepSzb srcWhole, int xoff, int yoff, float fx, float fy,
-//                     PtrStepSzb dst, int interpolation, cudaStream_t stream);
-// }
-
 }}}
 
 struct cv::gpu::SCascade::Fields
diff --git a/modules/objdetect/src/icf.cpp b/modules/objdetect/src/icf.cpp
index 416e5ffff0..9d0fb00dc6 100644
--- a/modules/objdetect/src/icf.cpp
+++ b/modules/objdetect/src/icf.cpp
@@ -40,7 +40,7 @@
 //
 //M*/
 
-#include <precomp.hpp>
+#include "precomp.hpp"
 
 cv::SCascade::Channels::Channels(int shr) : shrinkage(shr) {}
 
diff --git a/modules/objdetect/src/objdetect_init.cpp b/modules/objdetect/src/objdetect_init.cpp
index 77afeaa6e2..68e64243a9 100644
--- a/modules/objdetect/src/objdetect_init.cpp
+++ b/modules/objdetect/src/objdetect_init.cpp
@@ -40,7 +40,7 @@
 //
 //M*/
 
-#include <precomp.hpp>
+#include "precomp.hpp"
 
 namespace cv
 {
diff --git a/modules/objdetect/src/softcascade.cpp b/modules/objdetect/src/softcascade.cpp
index b34deeb60b..efebf33df5 100644
--- a/modules/objdetect/src/softcascade.cpp
+++ b/modules/objdetect/src/softcascade.cpp
@@ -40,7 +40,7 @@
 //
 //M*/
 
-#include <precomp.hpp>
+#include "precomp.hpp"
 
 namespace {
 
@@ -568,11 +568,11 @@ void cv::SCascade::detect(InputArray _image, InputArray _rois,  OutputArray _rec
     std::vector<Detection> objects;
     detect( _image, _rois, objects);
 
-    _rects.create(1, objects.size(), CV_32SC4);
+    _rects.create(1, (int)objects.size(), CV_32SC4);
     cv::Mat_<cv::Rect> rects = (cv::Mat_<cv::Rect>)_rects.getMat();
     cv::Rect* rectPtr = rects.ptr<cv::Rect>(0);
 
-    _confs.create(1, objects.size(), CV_32F);
+    _confs.create(1, (int)objects.size(), CV_32F);
     cv::Mat confs = _confs.getMat();
     float* confPtr = rects.ptr<float>(0);