Birds Eye View - Quick Start Guide

Getting Started/Evaluating the Product ⇨

This quickstart guide shows how to validate RidgeRun Bird's Eye View using sample images and a known calibration file. Use it to confirm that the BEV runtime starts correctly on NVIDIA Jetson, NXP i.MX8, or a Linux workstation before you move to custom calibration or application integration.

Prerequisites

You need:

An NVIDIA Jetson, NXP i.MX8 or x86 Linux machine.
Bird's Eye View source code (included with the purchase).

Install the Dependencies

The following dependencies need to be installed on the target machine:

Git: To clone the repository.
Meson: To configure the project.
Ninja: To build the project.
JsonCPP dev files: For the parameter loading.
OpenCV dev files: For image loading, processing and displaying.
GStreamer dev files and plugins: (optional) for image loading.
QT5 dev files: (optional) for image displaying.
CppUTest dev files: (optional) for unit testing.
Doxygen, Graphviz: (optional) for documentation generation.
Wget, Unzip: (optional) to download and unpack the sample images.

In Debian-based systems (like Ubuntu), you can run the following commands to install the dependencies:

# Update package sources
sudo apt update

# Install required dependencies with apt
sudo apt install -y \
  libjsoncpp-dev \
  libopencv-dev libopencv-core-dev \
  libopencv-video-dev libopencv-highgui-dev libopencv-videoio-dev \
  libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev \
  gstreamer1.0-plugins-bad \
  gstreamer1.0-plugins-good gstreamer1.0-plugins-base \
  gstreamer1.0-libav gstreamer1.0-plugins-ugly \
  qtbase5-dev qtmultimedia5-dev libqt5multimedia5-plugins \
  git wget unzip libcpputest-dev doxygen graphviz \
  python3-pip ninja-build

# Install latest version of meson with pip
sudo -H pip3 install meson

Build Bird's Eye View

Once you have dependencies installed, you can proceed to configure and build the project. Bird's Eye View can be built with or without CUDA acceleration depending on your hardware.

With CUDA Acceleration

Recommended for NVIDIA Jetson and x86 systems with NVIDIA GPU:

meson builddir
ninja -C builddir

Without CUDA Acceleration

On systems like the i.MX8 or x86 without NVIDIA GPUs:

meson builddir -Dnpp=disabled
ninja -C builddir

Run the Quickstart Sample

Once you've built the project, you can run the example. The project example already provides sample images and a calibration file.

cd samples && ./download_samples.sh && cd -
./builddir/examples/birds_eye_view -p examples/birds_eye_view.json

You should see the resulting image as below. Press ESC to stop the execution.

Quick start result — Resulting image after running the quick start guide

Troubleshooting

If the sample does not run:

Verify your internet connection. The sample app will attempt to download the input images, and will fail if there is no internet connection.
Verify the calibration JSON path.
Confirm that display and graphics dependencies are available.

Next steps

Birds Eye View/Calibration Guide to create your own calibration
Birds Eye View/User Guide/Minimal Application to integrate the library
GstBEV Plugin to build a pipeline
Birds Eye View/Performance/PC or platform-specific performance pages to compare throughput

FAQ

Can I use the quickstart with my own camera images?: Not directly. The quickstart uses sample assets. For real camera layouts, follow the Birds Eye View/Calibration Guide.
What should I do after the quickstart works?: Either create your own calibration file, integrate the runtime into your application, or test a GStreamer pipeline depending on your project.

Contact Us

Need help validating BEV on your target? Contact RidgeRun and include your platform, OS version, camera count, and target framerate/resolution.

Home