Recently I have been talking with a guy from my home town who owns a local oyster company. Every couple weeks/month he has his workers boat out to the oyster beds during low tide and they are patches of 10 acres or so and they have to walk in the mud and manually survey the amount of growth in different sections of the oyster bed to see where they need to harvest the oysters at that particular time.
He has seen my flying my quadcopter before and is interested in maybe investing in a system that they can use to fly over the oyster beds and have images taken which can then be stitched together by means of software so they can have one large image to analyze for the growth patterns of the oysters.
I know there are roadblocks when it comes to running a business with the use of drones, but that is not my concern right now because this project is in the planning stages and I just want some opinions on if this would be possible.
So what do you guys think? Is this a project that seems doable?
1 vote for definitely doable! Good luck
I agree that this is very doable from a technology point of view. The most appropriate system components will depend on the actual flight profiles and required image resolution but I don't see anything so far that would be a technical roadblock using currently available equipment.
If you want to see an example of a system that would probably work for this (may not be the optimum choice from a cost point of view but you can see that the technology is available) take a look at the work at this website...
I am thinking one of the main problems is that there are hundreds of rows of oysters and they all look similar, so if I used a gopro and used stitching software to align the images into one big image, would the software be able to do so given the similarity in all the images?
Another thing is that they are going to want a system that can take off and land from a non-moving boat, so is the arducopter software able to do this with relatively tight accuracy?
It depends on the stitching software you use. An interesting one is VReveal Pro, which takes a video file and makes a still panorama. It might be interesting to see what it does to a video looking straight down. I will add that to my list of things to do when the snow melts.
A good pilot can land in a square meter.
Well the VReveal Pro thing sounds awesome because if I could just have a video converted to a still image that would mean I wouldn't have to have pictures that need to be overlapped, triggered at the right time, etc. But for the accuracy I know about manual landing accuracy's, but I was wondering about the auto land features and how accurate they are?
You may want to reconsider using a Go Pro.. they are not the best for photogrammetry and mapping.
I've never tried autoland.
Later today I may try something using VReveal. I will attach my GoPro to the car (shooting sideways) and drive down the street. If I am right, I will have a single panoramic image of a few blocks. I'll post the results here.
I am definitely not dead set on using a GoPro. I might be looking into near infrared cameras because it may be able to pick up the growth patters of the oysters better. What cameras would you recommend for photomapping?
The Canon Powershot sx230 or elph 300 modified by maxmax.com is pretty nice!
vReveal worked even better than I anticipated.
I attached my GoPro2 to my car and drove down the block at about 10 MPH. The drive lasted about 30-seconds and covered a couple of thousand feet. When I used vReveal (demo mode) to make a panorama image, I was really surprised with the quality. Note that the vReveal demo is limited to 480 pixels (vertical).
Here is the original GoPro footage:
Here is the vReveal paorama:
Here's the file if you want to download it: https://dl.dropbox.com/u/1421565/copter/GOPR0237_panorama.jpg
So extrapolating this experiment, I don't see why a GoPro looking straight down wouldn't provide the oyster farmer the survey he wants.
Well thank you for taking the time to do that, the results do look good!. Now my only other concern would be how to have like 10 linear images like this and then connecting them into one large rectangular image of the whole oyster bed area.