Madeira Beach, FL camera station

Background

This coastal video monitoring station was installed by the U.S. Geological Survey (USGS) to collect observations that will improve our understanding of coastal change.

• the camera station began operating continuously in January 2017
• video/images are collected for 17-minutes at the beginning of every daylight hour
• standard Argus image products are produced:
  • snapshot
  • time-exposure image
  • time-variance image
  • brightest image
  • darkest image
  • runup timestacks

---

Hardware

The video camera is mounted on the roof of a local 6-story hotel, and the computer is located in a temperature-controlled elevator-control room on the ground floor of the hotel. On the roof, there is an electrical outlet to power the camera and an ethernet port that connects to a port in the elevator-control room, and in the elevator-control room there is also a LAN connection.

• Video Camera: Prosilica GT 3400C GigE camera by Allied Vision Technologies
• Camera Lens: Fujinon 12.5mm
• Computer: Cappuccino Echo77FW PC running Windows 7
• Camera Mount: ballast weighted, non-penetrating mount with vertical mast, connected to horizontal arm supporting the camera within a Pelco weather-proof housing

Camera mounted on the roof of a hotel in Madeira Beach, FL.

---

Software

Computer and Camera Control:

• Vimba SDK for Allied Vision cameras
• GoToMyPC (for remote desktop control)
• Tardis (time synchronization program)

Image Collection:

• Streams7 by IO Industries
• Python script

Image Processing:

• batch script
• Matlab executable
• Windows Task Manager (for automation)

Data Transfer and File Cleanup:

• batch script
• R script
• Windows Task Manager (for automation)

Image Post-Processing:

• Matlab executable
• Windows Task Manager (for automation)

---

Workflow

1. Computer Set-up

• log-in as Administrator
• adjust computer settings:
  • do NOT allow to go to sleep
  • disable Auto Install of Windows Updates so computer does not automatically restart
  • enable auto-restart after system failure/power outage
• set computer time:
  • set time to GMT
  • set-up time synchronization with internet time server (run Tardis with Windows Task Manager)
• configure ethernet settings
  • enable Jumbo Packets on camera LAN connection
• set-up GoToMyPC for remote desktop control

2. Camera Set-up

• install Vimba SDK for Allied Vision cameras
• install and configure JAI GigE camera driver
  • in the Local Area Connection Properties, check JAI GigE Vision Filter Driver, and uncheck AVT GigE Vision Filter Driver
• install Wavelet compression filter for Bayered images
• connect camera to computer
• open Vimba Viewer and configure camera settings:
  • set Acquisition
    • set AcquisitionFrameRateAbs = 2.0 (Hz)
    • set Trigger = FixedRate
  • set Controls
    • set Exposure, Gain, White Balance = Continuous (until desired lighting is achieved), then = Off
  • set Image Format
    • set image width and height
    • set PixelFormat = RGB8Packed

3. Camera Calibration

• determine intrinsic parameters - lens distortion
  • see Cameras/Lenses section
• determine extrinsic parameters - rotation and translation
  • deploy, survey, and image-capture multiple Ground Control Points (GCPs)
  • survey camera position
  • use code to solve for camera geometry parameters (azimuth, tilt, roll)
  • use parameters to geo-rectify images

4. Image Collection

• install IO Industries Streams7
• run Streams7 as Administrator
  • increase memory in Streams Administrator to 3000MB
• create new Video Library
• add New Device
  • set Compression, Driver, etc. following user manual
  • save Device Configuration to file
• set Export settings
  • File Filter = JPEG
  • edit Timestamp Options
    • Overlay timestamp on image
    • Use full date
    • Invert timestamp text (change color to standout)
• run Python script to automatically collect images every daylight hour (based on defined time frame)

Example of Streams7 program automatically controlling the camera image collection.

5. Image Processing

• run batch file to:
  • check that image collection/export is complete
  • run Matlab .exe to create image products
  • write metadata to image products

6. Data Transfer and File Cleanup

• on the local computer, use Windows Task Manager to run a batch file (for every hourly collect) to:
  • push/upload image products to eFTP
  • create manifest of copy/transfer listing new image products, including filename, filesize, and md5 checksum
• on the server computer, use Windows Task Manager to run R script (for every hourly collect) to:
  • pull/download image products from eFTP to server
  • run checksum on recently downloaded image products
  • create "cleanup" .txt files with commands to delete successfully transferred files
  • push/uplooad "cleanup" .txt files to local computer
• on local computer, use Windows Task Manager to run a batch file (once each evening) to:
  • run "cleanup" .txt files to delete data/files from local computer
  • delete corresponding Video Log files
  • note: raw images are not stored forever because of data storage limits on the local computer and file transfer limits for pushing to the server

7. Image Post-Processing

• post hourly snapshot and time-exposure on website
• use Matlab .exe to create rectified image products
  • use intrinsic and extrinsic calibration parameters
  • save image products as netCDF