Update for Rerun 0.11

CMakeLists.txt

@@ -9,7 +9,7 @@ if(NOT DEFINED CMAKE_CXX_STANDARD)
 endif()
 
 include(FetchContent)
-FetchContent_Declare(rerun_sdk URL https://github.com/rerun-io/rerun/releases/download/0.10.1/rerun_cpp_sdk.zip)
+FetchContent_Declare(rerun_sdk URL https://build.rerun.io/commit/746dbf3/rerun_cpp_sdk.zip) # TODO: 2023-11-28. Update to latest commit.
 FetchContent_MakeAvailable(rerun_sdk)
 
 find_package(Eigen3 REQUIRED)

src/collection_adapters.hpp

@@ -0,0 +1,75 @@
+#pragma once
+
+#include <rerun.hpp>
+
+#include <eigen3/Eigen/Core>
+#include <opencv2/core.hpp>
+
+#include <cassert>
+
+// Adapters so we can log eigen vectors as rerun positions:
+template <>
+struct rerun::CollectionAdapter<rerun::Position3D, std::vector<Eigen::Vector3f>> {
+    /// Borrow for non-temporary.
+    Collection<rerun::Position3D> operator()(const std::vector<Eigen::Vector3f>& container) {
+        return Collection<rerun::Position3D>::borrow(container.data(), container.size());
+    }
+
+    // Do a full copy for temporaries (otherwise the data might be deleted when the temporary is destroyed).
+    Collection<rerun::Position3D> operator()(std::vector<Eigen::Vector3f>&& container) {
+        std::vector<rerun::Position3D> positions(container.size());
+        memcpy(positions.data(), container.data(), container.size() * sizeof(Eigen::Vector3f));
+        return Collection<rerun::Position3D>::take_ownership(std::move(positions));
+    }
+};
+
+// Adapters so we can log an eigen matrix as rerun positions:
+template <>
+struct rerun::CollectionAdapter<rerun::Position3D, Eigen::Matrix3Xf> {
+    // Sanity check that this is binary compatible.
+    static_assert(
+        sizeof(rerun::Position3D) ==
+        sizeof(Eigen::Matrix3Xf::Scalar) * Eigen::Matrix3Xf::RowsAtCompileTime
+    );
+
+    /// Borrow for non-temporary.
+    Collection<rerun::Position3D> operator()(const Eigen::Matrix3Xf& matrix) {
+        static_assert(alignof(rerun::Position3D) <= alignof(Eigen::Matrix3Xf::Scalar));
+        return Collection<rerun::Position3D>::borrow(
+            // Cast to void because otherwise Rerun will try to do above sanity checks with the wrong type (scalar).
+            reinterpret_cast<const void*>(matrix.data()),
+            matrix.cols()
+        );
+    }
+
+    // Do a full copy for temporaries (otherwise the data might be deleted when the temporary is destroyed).
+    Collection<rerun::Position3D> operator()(Eigen::Matrix3Xf&& matrix) {
+        std::vector<rerun::Position3D> positions(matrix.cols());
+        memcpy(positions.data(), matrix.data(), matrix.size() * sizeof(rerun::Position3D));
+        return Collection<rerun::Position3D>::take_ownership(std::move(positions));
+    }
+};
+
+// Adapters so we can borrow an OpenCV image easily into Rerun images without copying:
+template <>
+struct rerun::CollectionAdapter<uint8_t, cv::Mat> {
+    /// Borrow for non-temporary.
+    Collection<uint8_t> operator()(const cv::Mat& img) {
+        assert(
+            "OpenCV matrix was expected have bit depth CV_U8" && CV_MAT_DEPTH(img.type()) == CV_8U
+        );
+
+        return Collection<uint8_t>::borrow(img.data, img.total() * img.channels());
+    }
+
+    // Do a full copy for temporaries (otherwise the data might be deleted when the temporary is destroyed).
+    Collection<uint8_t> operator()(cv::Mat&& img) {
+        assert(
+            "OpenCV matrix was expected have bit depth CV_U8" && CV_MAT_DEPTH(img.type()) == CV_8U
+        );
+
+        std::vector<uint8_t> img_vec(img.total() * img.channels());
+        img_vec.assign(img.data, img.data + img.total() * img.channels());
+        return Collection<uint8_t>::take_ownership(std::move(img_vec));
+    }
+};

src/main.cpp

@@ -8,7 +8,7 @@
 #include <opencv2/imgproc.hpp>
 #include <rerun.hpp>
 
-#include "batch_adapters.hpp"
+#include "collection_adapters.hpp"
 
 std::vector<Eigen::Vector3f> generate_random_points_vector(int num_points) {
     std::vector<Eigen::Vector3f> points(num_points);
@@ -18,6 +18,10 @@ std::vector<Eigen::Vector3f> generate_random_points_vector(int num_points) {
     return points;
 }
 
+rerun::Collection<rerun::TensorDimension> tensor_shape(const cv::Mat& img) {
+    return {img.rows, img.cols, img.channels()};
+};
+
 int main() {
     const auto rec = rerun::RecordingStream("rerun_example_cpp");
     rec.spawn().exit_on_failure();
@@ -51,8 +55,8 @@ int main() {
     rec.log(
         "world/camera",
         rerun::Transform3D(
-            rerun::datatypes::Vec3D(camera_position.data()),
-            rerun::datatypes::Mat3x3(camera_orientation.data())
+            rerun::Vec3D(camera_position.data()),
+            rerun::Mat3x3(camera_orientation.data())
         )
     );
 
@@ -64,21 +68,15 @@ int main() {
         return 1;
     }
 
-    // Log image to Rerun
-    cv::cvtColor(img, img, cv::COLOR_BGR2RGB); // Rerun expects RGB format
-    // NOTE currently we need to construct a vector to log an image, this will change in the future
-    //  see https://github.com/rerun-io/rerun/issues/3794
-    std::vector<uint8_t> img_vec(img.total() * img.channels());
-    img_vec.assign(img.data, img.data + img.total() * img.channels());
-    rec.log(
-        "image",
-        rerun::Image(
-            {static_cast<size_t>(img.rows),
-             static_cast<size_t>(img.cols),
-             static_cast<size_t>(img.channels())},
-            std::move(img_vec)
-        )
-    );
+    // Rerun expects RGB format
+    cv::cvtColor(img, img, cv::COLOR_BGR2RGB);
+
+    // Log image to rerun using the tensor buffer adapter defined in `collection_adapters.hpp`.
+    rec.log("image0", rerun::Image(tensor_shape(img), rerun::TensorBuffer::u8(img)));
+
+    // Or by passing a pointer to the image data.
+    // The pointer cast here is redundant since `data` is already uint8_t in this case, but if you have e.g. a float image it may be necessary to cast to float*.
+    rec.log("image1", rerun::Image(tensor_shape(img), reinterpret_cast<const uint8_t*>(img.data)));
 
     return 0;
 }