I am working on putting a system together and my concern is that I may need to gimbal the camera which is a canon A810 to keep my shots pointing mostly at the ground. Has anyone else had experience with this, and if you have could you list the gimal you have used and your experience with it?
I was thinking of going with the IRIS but I am concerned with its weight causing low flight times vs the Y6 for example.
If you're doing mapping, you don't need a gimbal. The stitching software will compensate for any changes in camera attitude during flight. Yes, for duration the Y6 is a better choice.
Thanks for taking the time to reply Chris!
I would recommend a gimbal. It makes life a lot easier.
I use an S100 with several gimbals. Most are designed by myself or a friend and 3D printed.
In my experience when using a gimbal it requires less post photo verification (less deleting of bad images) than without a gimbal.
AlexMOS boards seem to be the easiest to calibrate.
Checkout www.thingiverse.com, many styles of gimbals listed.
If you don't have a 3D printer I may be able to print one for you to try.
My flat bottom gimbal for the S100 is large so it would require a large multirotor, say an s800 or above.
We are a commercial operator and do Ag work in Australia by hanging two cameras (one NIR one visible light) from the bottom of the 3DR Hex and Y6. We now have 5 Y6s and they perform well every day (thank you 3DR and DIYDrones).
I agree with Chris, you don't need a gimbal. Just make sure the camera(s) are really well isolated from vibration, choose a decent altitude, fly slowly in full sunlight, at midday and in calm or light wind conditions.
Ensure your shots have a 40-60% overlay (we use 60 north-south and 45 east-west) In the case of stronger winds we increase the overlap to compensate for wind gusts.
Post processing (stitching) is helped by being at a reasonable (high) altitude - say 100m, set your copter up so it flies the grid always facing in one direction (we use north).
Of course a gimbal will help, but for efficiency and simplicity, try the non-gimbaled approach first.
Nigel, how do you have your camera's mounted if you aren't using a gimbal? I have a 3DR X8 and have tried encasing a camera in several layers of expanded foam, which works okay, but it is not a very user friendly option. Also, how do you always point the copter in one direction? I didn't even know that was an option during flight planning? Thanks.
Hi there Rick
Our simplest solution for two cameras; A modified Go-pro mount. Drill out the centre of the three forks, screw through a thread that matches your camera mount, place rubber grommets on each side, screw on cameras. (See picture)
Add vibration dampening between the go pro mount and the copter body. Then use chdk and a USB remote shutter release either:
If using two cameras as above (the pink camera is NIR, silver is VIS) you can use the same remote shutter release by creating a Y cable to the two camera USB ports.
Point copter in one direction is not a flight planning option it is a config parameter; Simply put, play with WP_YAW_BEHAVIOR (change to 0). Take off and point copter to the north before engaging AUTO.
Thanks for the picture and info Nigel. Are there any jam nuts or anything that the camera's screw down tight against to prevent movement during flight? I have successfully used both an intervalometer script and also the trigger by distance feature in MP. I also have two camera's, one NIR and one RGB and was hoping to make a Y cable to trigger both cameras.
By "prevent movement during flight" do you mean vibration or swaying? You will see the small black phillips head on the side of the mount - We secure the GoPro mount to the vibration-dampening plate with velcro and bands.
Swaying: - the overlap and height of the copter make this a non-issue for us. There is heaps of sway due to copter movement, wind etc but it is adjusted for in stitching software.
Vibration: - we concentrate on reducing the vibration of the rig (props, motors, chassis rigidity). Once this is under control, we have no problems with the camera - although we use the fastest appropriate shutter speed.
Just a note: our imagery is not "high definition". We spent some time trying to get perfect resolution so "a farmer could see a blade of grass" but have discovered that this is simply not required by the market. We get excellent NGV, NDVI etc.. results from low-res VIS (RGB) and NIR images. Photographers would be horrified - but to us it isn't about photography. it about data collection.
Thanks for the information, it really helps. Can you share your battery size/ cell count? I am also interested in how you handle the vibration isolation for the cameras I had not previously considered carrying both visual and NIR cameras at the same time, but it makes a lot of sense. Any chance you can share a few more photos of your aircraft? My plan was to do just as you said (already doing it with another project) using CHDK and letting the autopilot trigger the cameras. Although I found out today that the pixhawk outputs 3.3v signals which will not trigger a chdk camera. I am working on a level shifting solution now and will post back how I accomplish it if you would like.
Love the camera mount Nigel
you do not really need a gimbal - at least for generating DSMs. However, the less image processing (blending etc.) is required the more reliable and comparable are your results. Depending on the zenith angle of the sun and your flight direction/yaw you may see differences in the exposure of the images of neighboring survey grid lines. The effect is much reduced when using a gimbal. This is a photo of my Nadir gimbal for two cameras (RGB + IR) in action: