I live on a small lake in NH with an invasive milfoil problem. We control the plants with SCUBA divers handpulling the weeds. Spotting weeds has been done either by observers on the surface, or by divers running search patterns. Water visibility tends to be poor, about eight feet.
I have been spending time thinking about automating the problem, and started to think about assembling a solution when I stumbled on bluerobotics. This is great as I have been thinking about playing with Arduino.
My needs are fairly simple:
Run a search pattern in a specified area
Grab video or stills and record
Use high precision GPS to record location concurrent with video
Return to base and upload images/locations to be reviewed at high speed by observer to mark plants.
Extra credit someday for computer recognition of plants
Max depth about 15 feet
What to buy? I see a 4-6 week lead time which would get me going at the end of our season.
Having built an autonomous survey vessel similar to your description, except it was designed for recording the depth below the keel, I can assure you that it was neither a simple or trivial task.
I would consider using a towed video camera as apposed to an autonomous underwater ROV. Just correct the GPS towing vessel coordinates for tow tether length and descent angle.
Yes, thanks. Read the article when it came out. I was not specific: we have variable milfoil, not Eurasian. Variable milfoil is not susceptible to the weevil, and I believe the followup on the tests for Eurasian was disappointing. Hand pulling has been our best option, unfortunately.
