Gaps in sequential video filename numbers exist. These videos should be used in conjunction with other data such as co-located photographs and the analyses of co-located sediment sample which provide sediment grain-size data (see shapefile in this data release 2012-024-FA_samples.shp). Beginning with sampling site 31, where the sediment sampler and the imagery frame were tethered together, the location of sediment samples, photographs and videos are often coincident as they were collected during the same deployment of the modified SEABOSS. A notable exceptions to this are at sample sites 31 and 36, where the sampler did not collect a successful sample during the original deployment so a sample was taken approximately 100 meter south (sample 31) and 30 meters west (sample 36) of where video and photos were collected.
This dataset includes riverbed video in mp4 format and a trackline shapefile of the location of the ship for the duration of the video during USGS survey 2012-024-FA aboard the R/V Rafael in the Connecticut River. Eighty-eight sampling sites were occupied within the study area using a modified SEABOSS. Video imagery from the water column and on deck or were otherwise not usable were deleted. A total of 68 video files are included in this dataset; due to the riverbed turbidity and extremely poor visibility, no imagery was collected at sites 2-21. Sediment samples were collected at those sites; please consult the sediment grain-size data associated with this survey. The photograph and video system was deployed again beginning with sampling station 22.
Videos were edited but many still contain a significant amount of footage that do not show the riverbed. Many videos show a significant amount of suspended sediment in the water column, especially when the sampler touched-down on the riverbed. Some videos also have footage (typically less than 10-15 seconds, if any, at the end of the clip) of the sampler being recovered and hung over the side of or placed on the deck of the R/V Rafael. Notable exceptions are video clips from stations 1, 38, 65 75, 77, 80 and 82 which all have more than 15 seconds out-of-water footage (up to 2 minutes for station 1; stations 65 and 75 have 60 and 30 seconds, respectively, out-of-water in the middle of the clips).
Source_Information:
Source_Citation:
Citation_Information:
Originator: U.S. Geological Survey
Publication_Date: Unpublished Material
Title: sea floor video and Navigation
Geospatial_Data_Presentation_Form: raster and ascii text digital data
Type_of_Source_Media: disc
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 20120921
Ending_Date: 20120923
Source_Currentness_Reference: ground condition
Source_Citation_Abbreviation: video files and text data
Source_Contribution:
The SEABed Observation and Sampling System (SEABOSS) was designed by the U.S. Geological Survey (USGS) for rapid and effective collection of seabed images and sediment samples in coastal regions. The observations from video and still cameras and the sediment data are used to explore the nature of the riverbed/seafloor and, in conjunction with high-resolution geophysical data, to make interpretive maps of sedimentary environments and validate the acoustic remote sensing data. This particular setup of the SEABOSS was ultra simplified using only a GoPro HD Hero2 camera, a black and white drop video camera with topside feed, lights and a depth sensor attached to a ultra-lightweight square frame. At station 1 and stations 22-30, this small, simplified SEABOSS frame with the GoPro and video was hand deployed from the starboard side of the R/V Rafael. At these sites, no sediment sample was collected, however visual observations made from the video and photos at sites 22-30 were made and are included in this sediment sample dataset. Due to the poor visibility, the imagery portion of the sampling was halted and at stations 2-21 only the modified Van Veen sampler was used to collect sediment samples (without imagery). Beginning with station 31, through the end of the end of the sampling survey, the imagery frame (with the GoPro and video cameras) was attached to the modified Van Veen grab sampler and deployed together from the starboard side of the R/V Rafael. This sampler with the combined capibility of collecting imagery and a sediment sample, a modified version of the original USGS SEABOSS, is known as the mircoSEABOSS or the SEABOSSpro or the GoProSEABOSS. At a typical sampling site, the vessel occupied one of the target stations and deployed this modified SEABOSS. Usually at the end of a short photograph/video survey at each sampling site the winch operator lowered the Van Veen sampler until it rested on the riverbed. When the system was then raised a sample was taken from the riverbed and the sampler recovered to the deck of the survey vessel where a subsample was taken for analysis at the USGS Sediment Laboratory in Wood Hole, MA. The camera time during this survey was set to UTC; calibration photographs with the navigation system indicate that the camera time was 23 seconds ahead of GPS times, therefore a time offset was applied during photo processing. The GoPro camera was set to take a photo every 2 seconds while deployed. The best photos were acquired when the sampler was briefly set down on the riverbed. Likewise, the best video was typically acquired when the sampler was briefly set down on the riverbed prior to taking a sediment sample.
Process_Step:
Process_Description:
Eighty-eight stations were occupied during USGS survey 2012-024-FA with a modified SEABOSS (Blackwood and Parolski, 2001), equipped with a GoPro HD Hero2 and a black and white dropcam video camera collecting photographs and video, respectively. A Van Veen grab sampler was deployed at each station. Riverbed videos were acquired at stations 1 and stations 22-88 (the photo/video system was not used at stations 2-21 due to extremely poor visibility). Video files were recorded topside on the survey vessel to Digital8 Video (8mm DV) tape and DVDs. The DVD recordings were incomplete and video clips from several stations were corrupt. The DV tape represent a complete set of the survey data and were therefore processed for this data release. The original video files alone do not represent spatial data, however, post-processing the video with the navigation data and survey notes makes the link to spatially locate the imagery. The link is possible because the start and end times of the video file can be matched with corresponding time in the navigation logs. This process step and all subsequent process steps were performed by the same person - Seth Ackerman
Process_Date: 2012
Source_Produced_Citation_Abbreviation: Video files
Process_Contact:
Contact_Information:
Contact_Person_Primary:
Contact_Person: Seth Ackerman
Contact_Organization: U.S. Geological Survey
Contact_Position: Geologist
Contact_Address:
Address_Type: mailing and physical address
Address: 384 Woods Hole Rd.
City: Woods Hole
State_or_Province: MA
Postal_Code: 02543-1598
Country: USA
Contact_Voice_Telephone: 508-548-8700 x2315
Contact_Facsimile_Telephone: 508-457-2310
Contact_Electronic_Mail_Address: sackerman@usgs.gov
Process_Step:
Process_Description:
Convert from original media and split into individual files:
The SEABOSS video imagery was originally recorded on Digital8 tapes. Each tape was played back and imported to a separate iMovie project (version iMovieHD6). Within iMovie, each individual SeaBOSS deployment was visually sseparated as its own video clip. In order to batch export all of the clips for a particular iMovie project a Perl script (parse_iMovie_XML_Mar2014.pl) was created which parsed the iMovie XML project file and created a shell script (do_convert_12024.sh). Each line of the do_convert_12024.sh script was an one-liner command to run the program FFMPEG with certain parameters and output a video file named by station number. For example...
> ffmpeg -ss 0:0:1.701 -i 12024_DV1.iMovieProject/Media/12024\ DV-01.1\ 01.dv -c:v libx264 -crf 22 -preset slow -b:a 192k -ac 2 -y -t 306.707 Station_1.m4v
> ffmpeg -ss 0:5:8.408 -i 12024_DV1.iMovieProject/Media/12024\ DV-01.1\ 01.dv -c:v libx264 -crf 22 -preset slow -b:a 192k -ac 2 -y -t 113.38 Station_22.m4v
> ffmpeg -ss 0:7:1.788 -i 12024_DV1.iMovieProject/Media/12024\ DV-01.1\ 01.dv -c:v libx264 -crf 22 -preset slow -b:a 192k -ac 2 -y -t 276.31 Station_23.m4v
where "-ss 0:7:1.788" means begin video at 7 minutes 1.788 seconds, "-i 12024_DV1.iMovieProject/Media/12024\ DV-01.1\ 01.dv" is the input video file, "-c:v libx264 -crf 22 -preset slow" are the video transcoding parameters (high-quality H.264 video using the encoder x264, constant frame rate of 22, slow speed transcoding for higher quality results), "-c:a libfaac -b:a 192k -ac 2" are the audio transcoding parameters (using the ACC encoding libraries, bitrate of 192 kilobits per second, in stereo layout), "-y" overwrites the output file if it exists, "-t 276.31" is the duration of the clip to transcode to the new file.
Process_Step:
Process_Description:
The do_convert script was run for each iMovie project creating individual video clips in m4v format; ffmpeg (version 3.1.2). The resulting video files were renamed by prepending the field activity identifier to the original video filename (e.g. 2012-024-FA_Station_22.m4v).
Source_Used_Citation_Abbreviation: DV video files within the iMovie Project
Process_Date: 201709
Source_Produced_Citation_Abbreviation: Processed m4v video files
Process_Step:
Process_Description:
Converted the m4v files to mp4 files and removed the audio (which was a chirpy-sounding encoded GPS signal, a backup to the traditional navigation log) using a shell script that called on FFMPEG (version 3.1.2) video processing software.
> for f in *.m4v; do v=`echo $f | awk '{split($0,a,"."); print a[1]}'`; echo ffmpeg -i $f -c:v copy -pix_fmt yuv420p -an $v".mp4" |sh; done
Process_Step:
Process_Description:
A text drift log that includes the video filename, the sampling station number, the UTC start time, UTC end time, date and duration of the video file was created in preparation for creating a trackline file to associate with each video clip.
Source_Used_Citation_Abbreviation: survey notes and logs
Process_Date: 201709
Source_Produced_Citation_Abbreviation: drift log text file
Process_Step:
Process_Description:
The raw navigation feed from the Hemisphere Differential Global Positioning System (DGPS) receiver was logged at a one-second interval using Microsoft HyperTerminal on a Dell Latitude D820 laptop computer during the sampling survey. Log files were recorded for each survey day (by Gregorian day). The navigation files were processed using an AWK script (no version) called awkseth.gpgga.12024.awk by survey day and concatenated in to a single navigation file. Then a PYTHON script (SEABOSStools) was run to match up the video clips with the navigation log. (1) driftlog_gui_v4.py - PYTHON script that uses the drift log created in the previous process step and the processed navigation file described above. The output of this is the location shapefile for the riverbed videos with the fields: field activity identifier, sampling station name, year, date in the format JJJ (where JJJ is the Julian day (sequential day number starting at 1 Jan of the survey year)), video start time in the format HH:MM:SS where (HH:MM:SS is UTC time in hours:minutes:seconds), video end time in the format HH:MM:SS, GPS start time in the format HH:MM:SS and GPS end time in the format HH:MM:SS. The difference between the video start/end time and the GPS start/end time would be if the navigation was not collected at a 1-second interval. For this survey the video and GPS start were identical, and the same was true for the video and GPS end times.
Source_Used_Citation_Abbreviation: RAW navigation log files
Source_Used_Citation_Abbreviation: drift log text file
Process_Date: 201709
Source_Produced_Citation_Abbreviation: Sea floor video location shapefile (polyline)
Process_Step:
Process_Description:
XTools Pro (version 12.0) for ArcGIS (version 10.3.1) was used to reorganize, remove (video start and end times) and add new fields (TABLE OPERATIONS - TABLE RESTRUCTURE) to the polyline shapefile including adding an attribute for the video name (VIDNAME), survey ID (SURVEYID), device used to collect the imagery and samples (DEVICEID), the survey vessel (VEHICLEID), the date of imagery collection (DATE) and the Camera used (CAMERA). The length field ('Length_M') was populated using 'Calculate Geometry' (Property=Length; Use coordinate system of the data frame=WGS 1984 UTM Zone 18N; Unit=meters), which can be accessed by right-clicking on the attribute field name in the table view.
Process_Date: 201710
Process_Step:
Process_Description:
Added keywords section with USGS persistent identifier as theme keyword.
Process_Date: 20200908
Process_Contact:
Contact_Information:
Contact_Organization_Primary:
Contact_Organization: U.S. Geological Survey
Contact_Person: VeeAnn A. Cross
Contact_Position: Marine Geologist
Contact_Address:
Address_Type: Mailing and Physical
Address: 384 Woods Hole Road
City: Woods Hole
State_or_Province: MA
Postal_Code: 02543-1598
Contact_Voice_Telephone: 508-548-8700 x2251
Contact_Facsimile_Telephone: 508-457-2310
Contact_Electronic_Mail_Address: vatnipp@usgs.gov
