diff --git a/applications/gui/src/camera.cpp b/applications/gui/src/camera.cpp
index 48858d57fb74273d712a0c2e0ef48d87197e6d31..5a0964ab6e44f3a74ed5b61fbf799ca3b4226fda 100644
--- a/applications/gui/src/camera.cpp
+++ b/applications/gui/src/camera.cpp
@@ -11,6 +11,7 @@
#include <ftl/operators/antialiasing.hpp>
#include <ftl/cuda/normals.hpp>
#include <ftl/render/colouriser.hpp>
+#include <ftl/cuda/transform.hpp>
#include <ftl/codecs/faces.hpp>
@@ -134,6 +135,7 @@ void ftl::gui::Camera::draw(std::vector<ftl::rgbd::FrameSet*> &fss) {
if (fid_ != 255) {
for (auto *fs : fss) {
if (!usesFrameset(fs->id)) continue;
+ UNIQUE_LOCK(fs->mtx, lk);
ftl::rgbd::Frame *frame = nullptr;
@@ -147,6 +149,7 @@ void ftl::gui::Camera::draw(std::vector<ftl::rgbd::FrameSet*> &fss) {
texture1_.make(buf.width(), buf.height());
auto dst1 = texture1_.map(0);
cudaMemcpy2D(dst1.data, dst1.step1(), buf.devicePtr(), buf.pitch(), buf.width()*4, buf.height(), cudaMemcpyDeviceToDevice);
+ ftl::cuda::flip<uchar4>(dst1, 0);
texture1_.unmap(0);
width_ = texture1_.width();
diff --git a/applications/gui/src/screen.cpp b/applications/gui/src/screen.cpp
index 17f3291ef9d87f1ea5ac92307ab207739ee75e12..187ff7945d608f72ef45f4674101209bb3e2f7f7 100644
--- a/applications/gui/src/screen.cpp
+++ b/applications/gui/src/screen.cpp
@@ -38,7 +38,7 @@ namespace {
in vec2 vertex;
out vec2 uv;
void main() {
- uv = vec2(vertex.x, 1.0 - vertex.y);
+ uv = vec2(vertex.x, vertex.y);
vec2 scaledVertex = (vertex * scaleFactor) + position;
gl_Position = vec4(2.0*scaledVertex.x - 1.0,
2.0*scaledVertex.y - 1.0,
diff --git a/applications/gui/src/src_window.cpp b/applications/gui/src/src_window.cpp
index be7657c8c5c1d56ec2207aff2b02e7dcbbebee09..15e9e96f75f5b51c9e9a10f9004d3d597dd9974e 100644
--- a/applications/gui/src/src_window.cpp
+++ b/applications/gui/src/src_window.cpp
@@ -387,11 +387,13 @@ void SourceWindow::draw(NVGcontext *ctx) {
if (!camera.second.thumbview) camera.second.thumbview = new ftl::gui::ThumbView(ipanel_, screen_, cam);
camera.second.thumbview->setFixedSize(nanogui::Vector2i(320,180));
+ auto *iv = dynamic_cast<nanogui::ImageView*>(camera.second.thumbview);
+
/*if ((size_t)ipanel_->childCount() < i+1) {
new ftl::gui::ThumbView(ipanel_, screen_, cam);
}*/
//if (thumbs_[i].isValid()) dynamic_cast<nanogui::ImageView*>(camera.second.thumbview)->bindImage(thumbs_[i].texture());
- if (cam->getLeft().isValid()) dynamic_cast<nanogui::ImageView*>(camera.second.thumbview)->bindImage(cam->getLeft().texture());
+ if (cam->getLeft().isValid()) iv->bindImage(cam->getLeft().texture());
++i;
}
diff --git a/components/renderers/cpp/include/ftl/cuda/transform.hpp b/components/renderers/cpp/include/ftl/cuda/transform.hpp
new file mode 100644
index 0000000000000000000000000000000000000000..3044342b0e33a47b7d4784b307a27b04f4d75cf0
--- /dev/null
+++ b/components/renderers/cpp/include/ftl/cuda/transform.hpp
@@ -0,0 +1,23 @@
+#ifndef _FTL_CUDA_TRANSFORM_HPP_
+#define _FTL_CUDA_TRANSFORM_HPP_
+
+#include <ftl/cuda_common.hpp>
+
+namespace ftl {
+namespace cuda {
+
+ template <typename T>
+ void flip(
+ ftl::cuda::TextureObject<T> &out,
+ cudaStream_t stream);
+
+
+ template <typename T>
+ void flip(
+ cv::cuda::GpuMat &out,
+ cudaStream_t stream);
+
+}
+}
+
+#endif
diff --git a/components/renderers/cpp/src/CUDARender.cpp b/components/renderers/cpp/src/CUDARender.cpp
index 3affec4bdf172f61b88e7c24824996ac2e7fcd9b..6d0deb848509996105a550abc5197805b4939d8c 100644
--- a/components/renderers/cpp/src/CUDARender.cpp
+++ b/components/renderers/cpp/src/CUDARender.cpp
@@ -5,6 +5,7 @@
#include <ftl/cuda/normals.hpp>
#include <ftl/operators/mask_cuda.hpp>
#include <ftl/render/colouriser.hpp>
+#include <ftl/cuda/transform.hpp>
#include "colour_cuda.hpp"
@@ -538,6 +539,7 @@ void CUDARender::_endSubmit() {
accum_,
out_->getTexture<uchar4>(out_chan_),
contrib_,
+ false, // Flip
stream_
);
}
@@ -545,6 +547,10 @@ void CUDARender::_endSubmit() {
void CUDARender::_end() {
_postprocessColours(*out_);
+ // Final OpenGL flip
+ ftl::cuda::flip(out_->getTexture<uchar4>(out_chan_), stream_);
+ ftl::cuda::flip(out_->getTexture<float>(_getDepthChannel()), stream_);
+
cudaSafeCall(cudaStreamSynchronize(stream_));
}
diff --git a/components/renderers/cpp/src/colour_util.cu b/components/renderers/cpp/src/colour_util.cu
index bc8c1603398da9e3f94bad6e55b7d0dc3e3a09a0..b40ae62d6f42ed699c1b78beb88121b72c68ccf7 100644
--- a/components/renderers/cpp/src/colour_util.cu
+++ b/components/renderers/cpp/src/colour_util.cu
@@ -1,4 +1,5 @@
#include "colour_cuda.hpp"
+#include <ftl/cuda/transform.hpp>
using ftl::cuda::TextureObject;
@@ -156,3 +157,63 @@ void ftl::cuda::composite(
out.width(), out.height());
cudaSafeCall( cudaGetLastError() );
}
+
+
+// ==== Flipping ===============================================================
+
+template <typename T>
+__global__ void flip_kernel(
+ T* __restrict__ img,
+ int pitch,
+ int width, int height) {
+
+ for (STRIDE_Y(y, height/2)) {
+ for (STRIDE_X(x, width)) {
+ const T c1 = img[x+y*pitch];
+ const T c2 = img[x+(height-y-2)*pitch];
+
+ img[x+y*pitch] = c2;
+ img[x+(height-y-2)*pitch] = c1;
+ }
+ }
+}
+
+template <typename T>
+void ftl::cuda::flip(
+ TextureObject<T> &img,
+ cudaStream_t stream) {
+
+ static constexpr int THREADS_X = 32;
+ static constexpr int THREADS_Y = 8;
+
+ const dim3 gridSize(6,64);
+ const dim3 blockSize(THREADS_X, THREADS_Y);
+
+ flip_kernel<T><<<gridSize, blockSize, 0, stream>>>(
+ img.devicePtr(), img.pixelPitch(),
+ img.width(), img.height());
+ cudaSafeCall( cudaGetLastError() );
+}
+
+template void ftl::cuda::flip<float>(TextureObject<float> &,cudaStream_t stream);
+template void ftl::cuda::flip<uchar4>(TextureObject<uchar4> &,cudaStream_t stream);
+
+template <typename T>
+void ftl::cuda::flip(
+ cv::cuda::GpuMat &img,
+ cudaStream_t stream) {
+
+ static constexpr int THREADS_X = 32;
+ static constexpr int THREADS_Y = 8;
+
+ const dim3 gridSize(6,64);
+ const dim3 blockSize(THREADS_X, THREADS_Y);
+
+ flip_kernel<T><<<gridSize, blockSize, 0, stream>>>(
+ (T*)img.data, img.step/sizeof(T),
+ img.cols, img.rows);
+ cudaSafeCall( cudaGetLastError() );
+}
+
+template void ftl::cuda::flip<float>(cv::cuda::GpuMat &,cudaStream_t stream);
+template void ftl::cuda::flip<uchar4>(cv::cuda::GpuMat &,cudaStream_t stream);
diff --git a/components/renderers/cpp/src/dibr.cu b/components/renderers/cpp/src/dibr.cu
index 7530f0a4717d99a9a9ec62944c814ac37e7a9bff..2aa5987371afc61348ffe94f4fe0fc41929e4cb6 100644
--- a/components/renderers/cpp/src/dibr.cu
+++ b/components/renderers/cpp/src/dibr.cu
@@ -81,7 +81,7 @@ void ftl::cuda::dibr_merge(TextureObject<int> &depth_out, const float4x4 &transf
// ==== Normalize ==============================================================
-template <typename A, typename B>
+template <typename A, typename B, bool FLIPY>
__global__ void dibr_normalise_kernel(
TextureObject<A> in,
TextureObject<B> out,
@@ -97,24 +97,29 @@ __global__ void dibr_normalise_kernel(
//out(x,y) = (contrib == 0.0f) ? make<B>(a) : make<B>(a / contrib);
if (contrib > 0.0f) {
- out(x,y) = make<B>(a / contrib);
+ if (FLIPY) out(x,out.height()-y-1) = make<B>(a / contrib);
+ else out(x,y) = make<B>(a / contrib);
//normals(x,y) = normal / contrib;
}
}
}
template <typename A, typename B>
-void ftl::cuda::dibr_normalise(TextureObject<A> &in, TextureObject<B> &out, TextureObject<int> &contribs, cudaStream_t stream) {
+void ftl::cuda::dibr_normalise(TextureObject<A> &in, TextureObject<B> &out, TextureObject<int> &contribs, bool flip, cudaStream_t stream) {
const dim3 gridSize((in.width() + T_PER_BLOCK - 1)/T_PER_BLOCK, (in.height() + T_PER_BLOCK - 1)/T_PER_BLOCK);
const dim3 blockSize(T_PER_BLOCK, T_PER_BLOCK);
- dibr_normalise_kernel<<<gridSize, blockSize, 0, stream>>>(in, out, contribs);
+ if (flip) {
+ dibr_normalise_kernel<A,B,true><<<gridSize, blockSize, 0, stream>>>(in, out, contribs);
+ } else {
+ dibr_normalise_kernel<A,B,false><<<gridSize, blockSize, 0, stream>>>(in, out, contribs);
+ }
cudaSafeCall( cudaGetLastError() );
}
-template void ftl::cuda::dibr_normalise<float4,uchar4>(TextureObject<float4> &in, TextureObject<uchar4> &out, TextureObject<int> &contribs, cudaStream_t stream);
-template void ftl::cuda::dibr_normalise<float,float>(TextureObject<float> &in, TextureObject<float> &out, TextureObject<int> &contribs, cudaStream_t stream);
-template void ftl::cuda::dibr_normalise<float4,float4>(TextureObject<float4> &in, TextureObject<float4> &out, TextureObject<int> &contribs, cudaStream_t stream);
+template void ftl::cuda::dibr_normalise<float4,uchar4>(TextureObject<float4> &in, TextureObject<uchar4> &out, TextureObject<int> &contribs, bool, cudaStream_t stream);
+template void ftl::cuda::dibr_normalise<float,float>(TextureObject<float> &in, TextureObject<float> &out, TextureObject<int> &contribs, bool, cudaStream_t stream);
+template void ftl::cuda::dibr_normalise<float4,float4>(TextureObject<float4> &in, TextureObject<float4> &out, TextureObject<int> &contribs, bool, cudaStream_t stream);
// Float version
diff --git a/components/renderers/cpp/src/overlay.cpp b/components/renderers/cpp/src/overlay.cpp
index d1d83c9da5b19249029a7cf2d713890840ec4265..78415f5634f9ad90f695503adf1a5a3057f36856 100644
--- a/components/renderers/cpp/src/overlay.cpp
+++ b/components/renderers/cpp/src/overlay.cpp
@@ -288,7 +288,7 @@ void Overlay::draw(ftl::rgbd::FrameSet &fs, ftl::rgbd::FrameState &state, const
if (value("show_poses", false)) {
for (size_t i=0; i<fs.frames.size(); ++i) {
- auto pose = fs.frames[i].getPose().inverse(); //.inverse() * state.getPose();
+ auto pose = fs.frames[i].getPose(); //.inverse() * state.getPose();
auto name = fs.frames[i].get<std::string>("name");
_drawOutlinedShape(Shape::CAMERA, state.getPose().inverse() * pose, Eigen::Vector3f(0.2f,0.2f,0.2f), make_uchar4(255,0,0,80), make_uchar4(255,0,0,255));
diff --git a/components/renderers/cpp/src/splatter_cuda.hpp b/components/renderers/cpp/src/splatter_cuda.hpp
index da47b6fd3b2ef8ae0809728d27d876338207350f..e1021651ce1989cbf06b0a4b00fc04a1d457b844 100644
--- a/components/renderers/cpp/src/splatter_cuda.hpp
+++ b/components/renderers/cpp/src/splatter_cuda.hpp
@@ -141,6 +141,7 @@ namespace cuda {
ftl::cuda::TextureObject<A> &in,
ftl::cuda::TextureObject<B> &out,
ftl::cuda::TextureObject<int> &contribs,
+ bool flipy,
cudaStream_t stream);
template <typename A, typename B>