Phoenix Aerial Systems' Modular Software Suite
Our software suite provides a tight integration of all equipment during data acquisition, as well as a seamless workflow for post-processing.
To allow mapping in different scenarios, we provide the most flexible scanning architecture in the industry. To achieve this, the system is divided into the following components:
The Rover program runs on an embedded computer in the mobile mapping system. The software initializes, controls and monitors all sensors, logging all incoming data to a shock-resistant solid-state hard-drive. It also fuses data from navigation system, LIDAR and camera-sensors in real-time and streams the resulting point-cloud to the user's computer.
Rover can receive differential corrections from a GNSS reference station or public correction network, enabling accurate RTK-positioning. For most applications, this means that no GNSS/Inertial post-processing is required to generate accurate point clouds.
Of course, all LiDAR sensors shown on the LIDAR comparison page and all IMUs shown on the IMU comparison page are supported. New in version 2 is that up to 4 LiDAR sensors of any kind can be used at the same time. Another improvement is that more cameras are now supported: Many still cameras with USB connection are supported, allowing previews of the captured images to be sent to the ground in real-time, visualized in the exact position and orientation they were captured in. To check for image overlap and quality at the same time, preview sizes are configurable and can be sent alternating between downscaled and 100%-crop versions.
SpatialExplorer is the operator's tool to monitor and control mapping, running on a notebook with NVIDIA graphics and 64bit editions of Windows 7,8, or 10. As an industry-first, SpatialExplorer can render a real-time, three-dimensional representation of the point cloud during flight with low latency. In fact, this works so well that some pilots use this visualization for first-person-view (FPV) flying.
Next to the point cloud, the user interface of SpatialExplorer also presents an immediate summary of the system's status while scanning (see image) and can be customized to view other useful information such as satellite tracking and real-time vibration data.
Missions can later be replayed in time-lapse, real-time or slow motion, which can be used to analyze data or create videos. This is another feature unique to Phoenix Aerial System's software suite.
For both real-time display and in post-processing, it is possible to select only a subset of lasers, downsample the angular resolution, limit the sensor's range (both near and far planes), block subsets of the scanner's field of view etc.
All modules support extracting up to 15 returns (2 for Velodyne, 3 for Ibeo, variable for Riegl). Using a slider, only points from a selected range of returns can be displayed in SpatialExplorer. This feature allows the operator to more easily check in real-time whether the LiDAR is able to penetrate a forest's canopy and sample its ground.
After mapping, data from all sensors (navigatino system, LiDAR, photogrammetry) is downloaded from the rover using gigabit ethernet, which takes about 5% of the mission's duration. In the vast majority of cases, the deliverable will be a LAS file. To create it, SpatialFuser opens the rover's trajectory and LiDAR files, and fuses them into a georeferenced point cloud.
For aerial scans and most ground scans, the RTK-trajectory can be fused directly with the LiDAR data. When differential corrections were not available or satellite visibility was poor in challenging environments, GNSS/IMU-post-processing is recommended to create an accurate trajectory.
SpatialFuser can import trajectories directly from the Navigation System or from NovAtel's Inertial Explorer™ and thus, can export precise point clouds even of challenging environments.
By graphing all important navigation system parameters over time along with LiDAR coverage, the user can quickly get an overview of the mission. Analysis can be done by zooming into the graph, as well as seeking to the relevant time in SpatialExplorer and replaying the mission at the moment of interest.
Just like you can filter LiDAR data for the real-time view in SpatialExplorer, SpatialFuser is able to fuse LiDAR data only from a subset of lasers, with less angular resolution, or only from selected angular sections.
The resulting LAS files are industry standard and can be loaded into many GIS software packages (LAStools, QtModeler, OrbitGis, ENVI LiDAR, LP360 etc.). This allows for further analysis, like measuring distances, areas or volumes. Points can then be classified as ground, road, vegetation, architecture, vehicles or per your clients requirements.
For most mapping missions, it is desireable to send differential corrections from the GNSS reference station to the rover. To serve a customer base with a wide variety of applications, Phoenix Aerial Systems provides reliable transport of corrections in all imaginable scenarios. SpatialLightHouse is used to connect to the GNSS reference station, either using direct USB/serial cables when the reference station is in the field, or TCP/IP when it is not. After connection, SpatialLightHouse saves the station's raw observations to disk and forwards corrections to the rover in real-time.
SpatialLightHouse can run on the same computer that is used by the operator to run SpatialExplorer. Also, it can be run on a separate, small, cheap and low-powered computer. This way, the operator can use SpatialExplorer and move away from the reference station. This is helpful when the operator needs to travel together with the rover, e.g. when scanning from a car, boat or manned aircraft.
Corrections can be sent to the rover using multiple modes:
- a direct TCP/IP client-connection, when using WiFi and ethernet-based radiomodems.
- using a TCP-server: when the reference station is permanently installed in a fixed location, SpatialLightHouse can act as a correction-server on a machine with a public IP-address.
- using a serial port: this is used to forward corrections into a radio modem at the base, and requires a radio-modem receiver integrated into the rover
- without SpatialLightHouse: the reference station is configured to emit corrections on its serial ports, so a radio-modem can be connected directly
- using Phoenix Aerial System's connection-service: when GPRS/3G/4G service is available, rover, SpatialLightHouse (and SpatialExplorer) can connect to a common server and create a session, exchanging data in real-time through the cloud. Contrary to what might be expected, this has proven to be an extremely convenient and robust setup.
Instead of using a private GNSS reference station, SpatialLightHouse also supports connecting to differential correction networks using NTRIP. When such networks exist and are accessible while mapping, this can be a viable alternative.
In order to create an accurately georeferenced point cloud, the GNSS reference station must know the exact location of its antenna. Without this information, the point cloud will still be self-consistent, but might appear offset on earth. There are multiple ways of determining the reference station's antenna-position:
- setting up the antenna over a surveyed point
- using correction-services at the reference station (e.g. subscription-based satellite services via L-band, yielding an accuracy of about 6cm)
- by using the reference station to repeatedly measure and average its own position, with accuracy depending on the time spent measuring (example shown in image)
- by recording raw observations, and submitting them to user-positioning services like OPUS, AUSPOS, or equivalent. This will yield accuracies of about 3cm using raw observations as short as 15 minutes.
To use user positioning and other services, SpatialLightHouse can directly convert the raw observations to RINEX format.