EO-IR camera gimbal fitted to SurVoyeur mk-IIc - a new generation in Anti-poaching measures in Namibia. - Blogs

We have developed a version of the SurVoyeur mk-II, the 'c' version, which has a new nose housing a gimballed, stablized EO and IR camera combination.

This gimbal houses the FLIR QUARK 640 X 512 pixel Thermal camera, fitted with a shutter to enable flat field correction, and also houses the Sony EX11 Colour Block Camera.

Both camera are fully controllable from the ground station, colour modes, zoom where applicable, AGC, etc.

The gimbal is stabilized and fully steerable from the ground station. We are currently working on making accurate fitment jigs to bore-sight the cameras to the IMU, to enable accurate line of sight determination to allow automatic tracking of designated 'targets'.

Here are some photos of the various parts of the Gimbal, and the Gimbal fitted to the aircraft.

The two cameras with the QUARK shutter mechanism: The assembled Gimbal:

Pre-assembly:

A short video of the gimbal in motion:

https://www.youtube.com/watch?v=0xlB0HXT520&feature=youtu.be

Flight times are 70minutes at sea level 25deg C, 55 minutes at 1600m ASL, 40deg C.

Additionaly an 18mpixel Canon stills camera, stablised in roll, is fitted in the belly bay , under the wing, as was in the original mk-II aircraft.

search for SurVoyeur on DiyD for other blogs on mk-I and mk-II if interested...

The Nampilot.

We are still at our analysis stage and haven't yet purchased a camera or accompanying lens. For this purpose based on your comments, I tried preparing a MATLAB script (Considering the FLIR Quark 640, 19mm lens, 120m AGL and 20m/s flight speed) to get in some initial numbers. Could you please check it out and tell me if I'm on the right track and if possible suggest some improvements to make it more accurate. (link to script on google drive)

Also could you share some information on how to derive the processing power required for processing the footage from the cameras. Because I feel even if the cameras are capable of high frame rates and resolution we would be limited by what our onboard processor will be able to handle. This would also help me choose an optimum processor.

Thank You

You need to be a member of diydrones to add comments!

Join diydrones

Comments

Paul Meier January 15, 2015 at 9:27am

Brilliant !!!!!!!!!!!!!!
Divya January 15, 2015 at 9:09am

Hey nice gimbal!

Looks that you are using slip rings there? If possible can you share which ones are you using?

Also what is the weight of the gimbal including the sensor, a very nice build!
Gisela & Joe Noci January 12, 2015 at 1:12pm

Hi there Pransh,

All the casing parts are made from fibreglass/resin, using molds .

Regards

Joe
PRANSH RANA January 10, 2015 at 4:18am

what material have you used for the gimbal casing ?
geni December 12, 2014 at 9:36am

Hello Nikhil,

In search for a solutions for our UAV design, we found your great design.

We are building a nose rotating pod for our UAV, similar to yours.

I have a some questions about the roll mechanism:

In our design, the servo connected directly to the rotating part. Why did you put the pulley?

Also, we are looking for the best servos for that mechanism. We think about the Align DS610, do you have another recommendations?for the stabilization we are using the pixhawk card.

Thank You .
Nikhil Shetty April 17, 2014 at 5:51pm

We are still at our analysis stage and haven't yet purchased a camera or accompanying lens. For this purpose based on your comments, I tried preparing a MATLAB script (Considering the FLIR Quark 640, 19mm lens, 120m AGL and 20m/s flight speed) to get in some initial numbers. Could you please check it out and tell me if I'm on the right track and if possible suggest some improvements to make it more accurate. (link to script on google drive)

Also could you share some information on how to derive the processing power required for processing the footage from the cameras. Because I feel even if the cameras are capable of high frame rates and resolution we would be limited by what our onboard processor will be able to handle. This would also help me choose an optimum processor.

Thank You

Camera.m
Gisela & Joe Noci April 14, 2014 at 3:17am

Nikhil,

What camera are you using ( FLIR- TAU, Quark, etc) and what is the lens that you have fitted - From that info we can work out the field of view and therefore the ground coverage related to flight height AGL. This will give the ground movement rate per image pixel per unit time and then we can see at what speed the blur becomes unacceptable.

Regards

Joe
Nikhil Shetty April 13, 2014 at 1:46am

Thanks for the response.

We are trying to build a UAV and are trying to figure out what is the fastest we can fly to have unnoticeable motion blur based on IR camera resolution of 640x480, 120m AGL and 7-9 fps ("slow video" due to export restrictions)

Regards
Gisela & Joe Noci April 9, 2014 at 1:09pm

Hi Nikhil,

I am not to sure what effect you are referring to - there is certainly blur related to the camera frame rate and the rate at which the ground features pass in front of the camera, but flying at 18m/s and at 100meters AGL we do not notice the 'blur' at all. This type of artifact is related purely to the speed of movement of the ground features being imaged. At 30meters AGL and 18m/s flight speed, the image is quite un-viewable. However, a huge contributer to image blur is vibration, common on gas powered aircraft and multirotors. Since our plane is electric, single well balanced prop, we do not have a problem.

Regards

Joe
Nikhil Shetty April 8, 2014 at 1:15pm

I had a question regarding the functioning of the IR camera. I was wondering if you experienced any motion blur from the IR footage and if you did how did you rectify it.

Also would you happen to know what factors effect motion blur on IR cameras?

of 3

This reply was deleted.