Difference between revisions of "Eyesis4Pi 393"

Revision as of 10:52, 16 March 2017

Status

Available

Overview

Elphel Eyesis4Pi-26-393 is the newer generation of the full-sphere multi-camera systems for stereophotogrammetric applications. The system is calibrated to compensate for optical aberrations allowing to preserve full sensor resolution over the field of view. Distortion calibration is performed for precise pixel-mapping to automatically stitch images into panoramas, and also for photogrammetry and 3D reconstruction.   There are 24 x 5MPix sensors in the camera head, resulting in 120 MPix image acquisition. Panorama image resolution after stitching is 64 MPix, the equirectangular projection is 14000x7000. High-precision Inertial Measurement Unit IMU (ADIS 16375, 2400 samples per second) is integrated with the camera, as well as 5Hz GPS (Locosys LS20031) to provide precise position and orientation of the camera. Images, synchronized with IMU and GPS data, can be recorded to internal Solid State Drives (SSD), external swappable SSDs or to a PC over the network. Optionally Eyesis4Pi-393 camera system includes PC for image downloading and post-processing with pre-installed SDK and Free and Open Source Software for camera calibration, image acquisition and panorama stitching. Eyesis4Pi-393 can be mounted on a car for outdoor panoramic applications or carried by a person, which lets successfully create panoramas of small and narrow spaces.* Automated panoramic imagery and 3-dimensional panoramas are the 2 main applications of Eyesise4Pi-393 camera. * Car and backpack mounts are not included. Improvements over the earlier generation Eyesis4Pi new 10393 system boards compressor bandwidth: 1 GPix/s vs. 85 MPix/s only 3 system boards needed to support all 26 sensors vs. 9 contemporary OpenEmbedded Linux OS full 5 FPS at high quality vs. 2 FPS faster recording speed: 220 MB/s x 3 vs. 16 MB/s x 9 Gigabit Ethernet vs. 100 MB more compact more lightweight

Features and Specs

• 4*π (360x180) full sphere coverage
• 64 MPix panoramic image resolution after stitching
• Final panoramic image equirectangular projection dimensions - 14268x7135
• 0.05 Pix stitching precision
• Very small parallax - distance between entrance pupils: 46.5 mm
• High resolution lenses: 26x M12 lens mount
• Calibrated fixed lens to compensate for the lenses distortions
  • 10" (0.1 pix) in the center 80% x 80% area
  • 30" (0.3 pix) maximal error over the full FoV
  • <0.03um/° thermal expansion
• Precise image capture synchronization
• 5 FPS - maximum frame rate
• RAW or compressed image formats
• Images are geo-tagged (Exif) - 5Hz GPS receiver (Locosys LS20031)
• High-precision IMU ADIS 16375
• Recording to:
  • internal SSDs - not swappable - 3x m.2 SSDs (physical size: 2242 or 2260)
  • external SSDs - swappable - 3x 2.5" SSDs connected in a hot swap rack
  • PC over network
• Web Based GUI
• SYNC cable allows:
  • receive and log external trigger (from a sensor such as odometer or extra GNSS system)
  • generate output signal to trigger other sensors or cameras
• Weatherproof camera body and camera pole
• Dimensions: 230x760 mm (9x30")
• Weight: 8 kg (18 lbs)
• The Eyesis4Pi-393 system is highly customizable
• Free Software and Open Hardware

Image samples

• Post-processing results (equirectangular projections) and demos:
  • Samples page

Operation

• The system is controlled from a laptop via a web-based interface (Eyesis4Pi GUI). Which also displays real-time previews and parameters.
• Power consumption is about 50W.
• Power supply:
  • AC/DC power adapter (input: 110V/220V, output: 48V 2A) and can be used with a car power adapter (inverter), when mounted on a car roof.
  • 48V batteries (48V 4Ah - about 1-2 hours of operation)

Eyesis4Pi on a car	Fig.1 Cables
Fig.2 Recording to swappable external drives	Fig.3 Recording to internal SSDs or over network	Fig.4 Downloading data from internal SSDs to PC

Calibration

We have developed full calibration process and post-processing software to compensate for:

• Optical aberrations, allowing to preserve full sensor resolution over the camera FoV.
• Distortions – for precise pixel-mapping for photogrammetry and 3D reconstruction.

The calibration is performed using the machine (goniometer) that rotates the camera about 2 axes by precisely known angles. To calibrate IMU one needs to match the goniometer angles with the IMU data.

Check out these articles for more details:

• Building and calibrating Eyesis4Pi
• New calibration facility

Calibration process

Calibration process drawing

Software

All software is FOSS (Free and Open Source Software). It is experimental software that is used to test the camera performance, and it continues to be developed. All updates are available for download and build from GitHub. Software may require integration with customer software depending upon camera hardware configuration and application. It includes:

• Camera internal software/firmware and SDK
• Post-processing tools
• Calibration data

3D Model

Eyesis4PI-26-393 3D

More information

• Image sensors specs

Head Contents Schematic

FOV of sub-cameras in Eyesis4Pi-393, color shows time of acquisition (from red to blue):1-8 horizontal sensors; 1t-8t top sensors;1b-8b bottom sensors

How to buy

We are directly accepting orders and ship worldwide. Please contact Elphel for Lead Time and Quotes:

• Email (Sales Inquiry): sales@elphel.com
• Phone: (801) 783-5555 x 107
• Toll-free: (888) 3 ELPHEL
• Fax: (801) 812-8267