All Posts (14049)

Sort by
100KM

From Nothing To Landing

3689531994?profile=originalLow cost UAV has been advance in leap and bounce in recent year. Today general public can spend a few thousand dollar and a set of fully automatic UAV can be purchase and start taking aerial image and video. At least most of those who selling it hope you"ll to believe this is how simple is it.

Being involve in RC for the last twenty years, we use to spend a few months to help a new RC pilot to earn his wings. Along the way may be a few repair or sometime a few kits are require for the unlucky one.

Today technology has been greatly advance, we have flight simulator, autopilot and so on to help us to complete our flying mission. So is it true that now we can buy a UAV system off the shelf and start using it to get the data that we require ?

Recently I have an opportunity to evaluate if low cost UAV can be integrated to a GIS team of a plantation company. My project is to supply a low cost UAV system, train up a team of GIS executive staffs ( six of them ) to operate the UAV within three training session. Three days per session.

The objective is after the training session, the staffs should able to operate the UAV independently. Because of the restricted open space available in the operation area, and possible wind direction change, automatic landing feature in the autopilot may not a good choice. Manual landing seems mandatory in this situation. So manual landing is one of the our main objective.

The UAV system is Skywalker base with APM and Mission Planner as autopilot and ground control station.

3689532214?profile=originalThe training is divided into four sections. Basic radio control airplane operation, ground control station software training, Flight simulator training and real radio control airplane flight training.

The first two section is relatively easy, Basically the students are introduced with the basic knowledge of radio control airplane, radio control system and related electronic. For ground control station, the students were introduced with the various interface and function of Mission Planner, how to plan a mission, mission monitoring and so on.

The harder parts are the flight simulator and actual radio control airplane flight training. It is easy to say : up elevator.......turn left........level......ect but for a guy who never try radio control airplane, he really do not have the eye-hand coordination require to control the airplane in the air. Another problem is a new RC pilot not use to see the difference position of an airplane in the air.

To help the trainees gain eye-hand coordination as quick as possible, we use flight simulator. Our choice is Realflight 6. It is an excellent flight simulator with great graphic. We start the simulation with the students flying around freely. Very quickly they’ll find it is almost next to impossible to control the airplane to fly in the way they wish. After sometime, I fly a right hand traffic pattern as demo and ask them to do the same.

The simulation training session for each student is 10 minutes. I found that after 10 minutes or so, the student will start to lost concentration and subsequence training time is counter productive.

After a few sessions, all the students had develop some eye hand coordination and able to control the airplane to fly some recognisable traffic pattern but the altitude control is poor. In the evening, I let the students try their hand on my Skywalker with APM under stabilized mode. The reasons to let them fly real RC plane as soon as possible is let them had a feel of how the real things feel and looks while it was in the air. It is to my surprise that all of them can fly the right hand rectangle pattern at three mistake high with verbal instruction. At the end of the first training session, the confident of the students soar as they see for themselves that they can actually fly the plane with guidance.

The next training session is schedule two weeks after the first session. The students are require to practice computer flight simulation during their free time. These two weeks time is design to let the students develop eye-hand coordination. There is really no better way other than time to let the student to develop and acquire the connection and reflex require to fly RC plane.

We meet up two weeks later, now the students has trim down to 4 persons. Two of them really had a hard time to get the plane under control with flight simulator. It is true that everyone can learn to fly RC plane but some may need longer practice.

This session will emphasis on actual flight training with some knowledge of APM parameter and tuning. We go to the flying field as early as 3.00 pm. Arm with the experience acquired from the flight simulator, three of them can fly comfortably with some verbal instruction. Among the three, one of them excel in showing total control of the plane. After a few flight, I let him try some low pass and looks like he can handle it. Other students continue to fly traffic pattern in safe altitude.

In the late evening when the wind was calm, l take the risk and let the good student try to land the plane under stabilized mode. To my surprise he actually able to land it in one piece although he miss the runway. Subsequent flight show that he can land the plane consistently. In the following two days training, all the students can made a landing although some landing approach looks scary, but none of them broke the plane.

Now the training was completed and back to the mission field. Can the students handle the UAV as expected ? I wish to say under my high quality, super efficient training program......bra bra bra......mission accomplished.

Unfortunately the answer is mixed. All of them can plan a mission, set up the camera and assemble the plane ready for mission, but even the best student do not has confident to land the plane in the mission field. The landing area is a small football field surrounded with building and oil palm tree. Therefore in the initial stage, I send an experience RC pilot to help bring the plane down. ( you may guess from the picture who he is )

The students continue their training in the evening after mission flight. Two of them can actually make good landing in calm condition after a few days of hard work. But they still need more practice and experience to become a proficient pilot.

3689532200?profile=originalIn conclusion, I feel that with the help of computer flight simulator, students can master the basic flight technique very fast. But to be able to have total control and made consistence good landing approach and  eventually land the plane really need some time for the skill to develop. We need time to let our eye, brain and hand to communicate effectively and work in harmony.

Finally, Is low cost UAV system plug and play ? I'm sure not quite yet. For a new user to operate our low cost UAV, quality training is very important to ensure safety and effective UAV operation. Some potential user may miss lead by "autopilot" thinking that everything is automatic. I hope anyone who is planning to buy a UAV, regardless if it is low cost or not, quality training should be integrated in the purchase. This, in my opinion, is the most important step toward successful UAV operation.      



Read more…
3D Robotics

Farm drones are hot!

3689532145?profile=originalThis magazine isn't online, but you can get it on your iPad, Kindle or Android tablet via NextIssue. The article is a pretty basic overview, with examples of using a Rite Wing Zephyr and a hexacopter along with a Canon s100 modified with the IR filter removed. But it's notable that drones are getting this kind of attention in agriculture. 

Read more…

Another diversion this morning..

I had an extra camera, antenna and video xmitter so, here is what I did with it.

3689532133?profile=original

This is my Composite ARF "Spark".  It is an electric EDF aircraft, weighs about 13 lb's, has about 11 lb's of thrust, uses's a 10s setup and draws about 94 amps at full throttle. Its easy to fly and will go in excess of 150 MPH

I taped a camera, transmitter and antenna that I had "modularized" to the bottom and took it for a flight.   I just take this assembled set of electronics and tape it onto whatever I want to fly!

I use 5.8ghz and this is just a 100 mw xmitter.

The video below was made at my local flying field. It was quite windy and I bounced around a little bit on landing and it is obvious that the nose gear leg bent a little. After this flight, i just bent it back by hand and it was ready to fly again..

RE

Read more…

Hot in Texas

I took my Ardu copter out for a test flight to see if I have helped the stability with my latest attempt to tune some PID's.

It is definitely better, but I have an opportunity to make it better for sure. (I am slowly learning).

Anyway, I am gaining more confidence and so here is a video of a short FPV flight, recorded from my ground video station.

I enjoy looking at recorded video with the OSD info, I notice on this flight I get "Low RSSI"  continuously, yet I don't have a connection to the receiver, also, I get overspeed at times and I get stall warnings.. Also, I get a message "Low battery" even though I am at about 60% capacity. I guess I need to go into the setup on the Minim OSD and work on the settings!

Read more…

Arducopter crash. Throttle wire fell out?

3689532010?profile=originalI was flying my hexacopter at a local school on june 22. Everything started out great. I was flying it in auto, loiter, guided, and RTL mode. While i was flying my hexacopter started drifting away because of the wind. At witch point i brought it back and tried moving the throttle down to land it, but the throttle wasn't working. I decided to flip the switch and put it into RTL mode, after that the whole thing died, the motors lost power and fell out of the sky. Im not shure exactly what happend but i think the throttle wire fell out.
3689531962?profile=original3689532043?profile=original3689532077?profile=original

Read more…
MR60

X8 Octo Quad build log & maiden flight

3689531762?profile=original

Here is a short build log of how I converted my APM2.5 Quad to a X8 Octo quad.

First I dismounted everything which is always a pain, especially when you have to disconnect these little bastard DF13 connectors. By doing so a few times already (it is my fourth build with same APM...), I had finally to replace the telemetry cable by a new one. Oh well not too bad...

I could scrap the base 3DR power distribution board which was not meant for 8 ESCs and I got instead a vulcanUAV PDB which allows for (way too) many ESCs and 250 amps as seen on this picture.

3689531582?profile=original

I soldered also a 3DR power module and connected a Y battery splitter to be able to connect two batteries in parallel.

 

3689531784?profile=original

I prepared the motors. These are 3110-17 (700Kv) tiger motors. I love 'em. As you notice on the picture, I took my lessons from the first builds : every part is now tagged, either with a number, either with a color sticker, in order to know what motors goes on which arm and which cable connects to which ESCs (and so to keep the correct spin direction).

These motor cables are quite long so I decided to braid them, giving an extra EMI improvement and shortening them a bit:

3689531675?profile=original

By dismounting the quad, I noticed that the four screws holding the motors to their motor plate came a little bit loose, due to vibrations. I do not want on these parts to use blue loctite because it is a a nightmare to unscrew (for my next fifth build...). I used instead some hot glue as seen there:

3689531696?profile=original

(yeah, I know, I had to mix different screws, not pretty but who cares, nobody sees the bottom motor plate anyway...)

Then I laid on the floor the bottom plate, passed the braided motor cables in the arms, connected all the motor cables to the ESCs (thanks to the color stickers), placed the cables as neatly as possible. I shortened the ESCs power cables to the minimum in order to gain a clean placement. I soldered directly the ESCs to the PDB (no risk of bad connectors). By the way, a lesson I've learnt when you solder so much on a board : get a very high power solder iron otherwise the heat latency of the board gets unmanageable (and gets worse for every ESCs that is added).

I decided to keep only one BEC power cable from one of the ESCs. The other are snugged under the PDB (No, I did not cut them, you never know what will be needed in a next build)

 

3689531857?profile=original

3689531917?profile=original

Then it was time to work on the APM itself and all of its accesories : receiver, beeper, telemetry, GPS, etc. In order to get vibrations as low as possible, I tried what you see on the picture : APM is mounted on a small plate about 3cm above the center plate, with rubber dampers-white nylon spacer-silicon damper (what you see between the two metal washers).

This carbon fiber plate is thus never in direct contact with the rest of the frame. APM itself is placed on four corners of moongel (I put a balse plate between the carbon plate and APM to raise it a bit so that APM does not touch the metallic washers).

 3689531806?profile=original

APM is pressed gently on the moongel with two rubber bands. Later on, after I attached the GPS, I also improved a bit the vibrations levels by adding moongel between the GPS on top of APM and the APM case, so that the APM is sandwiched between two layers of moongel.

3689531938?profile=original

 This is how the Electronics look once finalized and all cabled.

Then I took the beast out for a short maiden flight as seen in the video herebelow:

 

 

 

 

Read more…

Laser cut frames my thoughts and for sale

So when i first got into copters, i saw i could but all sorts of frame, but to be honest, they were boring and more importantly it was not  challenge or any fun to just build a out of box frame, so i decided i would laser cut my own frames.

3689531885?profile=original

Now i will point out, i sell these frames, but i can also design a custom frame for you, the frames sell for $30 each

a custom frame costs $35.

I can also custom make other things for you that use the laser cutter.

The frames can also be made form various color acrylic if you do not want plywood.

they cost $40 as acrylic is far more expensive,

 

Contact me Jared_reabow@hotmail.co.uk

Skype ox141jf@hotmail.co.uk

 

 

 

 

 

 

 

 

Read more…

The End of Wiring Woes

3689531742?profile=originalHi everyone! I am finally ready to show off something I've been working on for months! It's a motherboard (which I am calling the MAXboard) that all the various electronics for a multirotor aircraft plug into. It eliminates pretty much any possible wiring mistakes, and all soldering. It makes it much easier to build, repair, and upgrade your aircraft.

This is an open source project, and am I running a Kickstarter campain to get some funds to continue development. I am also really hoping to get the support of the DIY Drones community to help me test it and work out bugs.

There is also a frame I designed to go with it as well as a ground station I think you'll like. I'll post more information on those, as well as some videos, soon.

For more info, or to help support the project, please take a look at my Kickstarter. Thanks! http://www.kickstarter.com/projects/1703258614/maxrotor-open-source-plug-and-play-modular-quadcop

Read more…

3689531828?profile=original

SO i have been on the lookout for a camera gimbal to lift some heft cameras over 2kg but everything i could find was at least $200 whilst the smaller gimbels for DSLR camera were a minimum of $150

This was frustrating for me, i could not understand why the prices were so exorbitant, this frustration was made even worse when i found out go pro gimbles came in at $70 whith really crappy ones being cheaper, so i decided to make and sell my own.

I will sell my predesigned one, or can make various different designs for smaller or bigger cameras and copter, with or without legs etc, to make a final design custom gimble i charge $5, a rough sketch is free.

So contact me to discuss custom gimbels or negotiate a purchase.

Contact me at Jared_reabow@homtail.co.uk

Skype on on ox141jf@hotmail.co.uk

3689531837?profile=original

This is what i came up with

I will design a mount plate to suit you, however you wish it to mount, you can with the 4 holes in the top that i have since added you can mount using heavy duty elastic bands to reduce vibration. or fix t directly to the frame with cable ties.

It is designed to work with or without servos as a free floating gimble which means it uses gravity to keep the camera level or uses servos which allow you to manually tilt the camera and stabilize with a stabilization board such as the apm

I will not include the servos so you can choose whtever servos you like, the current mount method for the servos has one slotting into the center as you can see in the first photo and the other mounted on top glue in place for added strength.

Both can use wire/string or control horns to connect the servos to the frame, it is entirely up to you guys.

IF you want me to mount the servos and add servos to the gimbal i will charge  $20 to make a total of $70

Contact me at Jared_reabow@homtail.co.uk

Skype on on ox141jf@hotmail.co.uk

Read more…
Moderator

Mugin 3M v III

The new design of the Mugin UAV 3M is out.

It's a huge UAV platform with 300 mm wingspan. The platform caters to larger UAS projects such as universities or commercial projects. It's really not a beginner platform.

 3689531709?profile=original

The platform is available here: Mugin UAV 3m vIII

It comes with or without engine/prop.

 

 

3689531654?profile=original

3689531561?profile=original

3689531723?profile=original

Twin Motor

3689531665?profile=original

Type: Two-stroke piston valve petrol aircraft engine
Displacement: 55cc(27.2ccx2)
Bore x Stroke: 34mm×30mm
Carburettor: Walbro(Diaphragm & Butterfly Valve)
Max. Output: 5.6ps/7600rpm
RPM Range: 1600-7800rpm
Weight: About 1780g(Engine+muffler+CDI)
Ignition power: Auto advanced CDI(DC 4.8 Volts , 4 cell NiCad or NiMh battery)
Fuel mixture(Ratio of gasoline and oil): 25-40 : 1
Propeller: 20X12-23X8 two-blade prop( suggest to use 22x8 prop)

Read more…

 

3689531551?profile=original

I am loving the APM 2.5+.  I am in the middle of a custom quadcopter build - but I wanted to spend some time getting used to the APM 2.5+ / Mission Planner.  So I pulled out my 6x6 Rover built on the MINDS-i robotic platform : MINDS-i.  I used my 3D printer to make a platform to hold the APM 2.5+ and some velcro / zip ties got everything wired up - you will notice I need to get some nice length female / female servo wires, I cheated and just use jumper wires for now.  Test #1 - I learned I didn't setup the input from my Spektrum DX6i Transmitter / Receiver correctly, so I after laying out the plan and heading out to this church reception parking lot - all I could do was run it around manual - at least the RC was working.  I reviewed the site and rewired.  Test #2 - I tried to do a small area around my house in my driveway,  didn't work so well, I think the area was too small, the house / garage likely blocked up the GPS and I noticed my Rover was drifting.  

Aligned up the steering, laid out a nice mission and went back out to the large parking lot.  Test #3 - third times a charm.  It ran through the waypoints, I did notice it didn't return home but stopped on Waypoint #4.  I was able to kick it into manual drove up facing away from Waypoint #1  - kicked auto is spun around and re-ran the route.

Meanwhile I picked up my AR Drone 2.0 to get some video of the rover.  I also have the GoPro on the Rover so I have some video to download from that.  I will post some pics and links to video on here.

3689531633?profile=original

 

 

3689531478?profile=original

 

Video Links

======================================================================

CDAutoRover Run 1

 

CDAutoRover Run 2

 

Check out my channel, I have some other robot goodies: RxDTxDTTL (Chuck D)

 

Read more…

3689531434?profile=original


Hello to all. This is my first post and I am hoping to raise some awareness of how UAV's can help in the fight against poaching. There is an organisation called the International Anti Poaching Foundation (IAPF) which is dedicated to stopping the slaughter of endangered species in Africa. Some of the videos on the website made me sick, to think poachers kill these magnificent animals just for their tusks and horns that are used in useless Chinese remedies is nothing short of a disgrace and travesty on the human race.

These guys are just starting to use drones to track poachers and they need our help. They currently have 2 in the air and are in the process of preparing another 5. This is their own UAV program and is fully funded by donations. If you are truly interested in taking your drone skills to the ultimate level then please do what you can to help. Make no mistake about it, this is a war they are fighting. Hunting and capturing poachers and protecting wildlife is a full time job. The organisation is run by an ex SAS (Special Forces) soldier from the Australian Army. Some of the accounts from engagements with poachers are terrifying and usually end in shoot outs no different to a warfare situation. Rangers have been killed trying to protect these animals. The 2 drones they do have have saved their lives many times.

As an ex Royal Australian Air Force missile engineer I am just starting to help them out and I will do whatever I can. But as a newbie to the drone community I am here to get information and toss some ideas around, raise awareness and also see what equipment people are willing to donate or to help out in whatever way they can. There are 39000 members of this community, if 1 in 1000 can help out even in the smallest way you can contribute, the difference you will make will be greatly appreciated.

Flying a drone around the neighbourhood is fun but to lend our skills to this, to me is the ultimate use of this technology and my years of training in guided missile electronics and control systems. Please contribute anyway you can.

So its over to you guys and lets see what we can do.

Comments please

Best Regards

Michael

Read more…

Five seniors at Portland State University recently finished a 6 month capstone project:  to develop a computer-vision system for a quadcopter.  As you can see from the video, our vision system tracks a fluorescent pink circle on the ground and tries to maintain a stable hover over that target.  The code can be easily changed to track virtually any color, however.  The system runs on a Raspberry Pi attached to the copter, has one webcam pointing straight down (with a fisheye lens for maximal viewing angle), and when Raspbian boots it launches MAVProxy, an open-source Python mavlink wrapper.  The computer vision code is just a module that gets imported into the MAVProxy environment.  The copter takes autonomous control if the copter is placed into the ALT_HOLD flightmode with the RC controller AND the target is present in the frame.    

Our code can be found on Git at dwrtz/MAVProxy.  For those of you already familiar with MAVProxy, you'll want to grab mavproxy_vision.py from the 'modules' folder.  The README.vision file in the root of the repo contains tips for configuring your vision tracking system, including the code dependencies and instructions on how to use the stuff in /modules/vision_utils.  We hope that drone enthusiasts with Python skills will take our code and improve upon it.

The git repo will undergo improvements in the coming weeks, including a copy of the technical report and links to all the parts we used.  Until then, any questions can be directed at colindoolittle@gmail.com.  Replies may be delayed in the next week as I'm going on vacation, but will be answered eventually.

Thanks! 

Read more…
Developer

[Update] DroidPlanner v0.9.0

Another update on DroidPlanner, now it has nicer HUD. The new app (v0.9.0) is on Google Play. 663c4da8-d850-11e2-992b-7d5adc28ff08.png

The new version has the following updates ( v0.9.0 over v0.8.0):

  • New HUD
  • Removing old Terminal screen
  • APM level calibration using a menu button.
  • Ability to read/save parameters from/to a .param file.
  • Russian language available
  • Greek language available
  • Latvian language available
  • German language available
  • Fixed some bugs
  • Improvements to the code structure

I would like to thanks everyone that is getting involved with the project, developing, donating and reporting bugs. Here is a list of the developers. I have made a small video of the project history (using gource), here it is (run in 720p to see the people/folders names):

For more information I suggest the DroidPlanner Wiki page. And as always if you want to help consider donating for the project by buying this app, joining the development team, or reporting a issue/bug/improvement on GitHub.

Read more…
3D Robotics

3689531455?profile=originalGood interview with ArduCopter flier John Vigg in Vice Magazine's The Creators Project. Excerpt:

John Vigg thinks everyone should have a personal drone. Maybe you’d use yours to uphold national security, or maybe you’d have tacos delivered to your apartment. Whatever. 

While people have been strapping high-def cameras to their heads and helmets for years now, Vigg still thinks they’re missing the point. He’s going for a new aesthetic. In the search for this “new aesthetic,” Vigg has designed and programmed drones armed with cameras to fly over various areas that are otherwise unreachable by foot. 

His interest was piqued by a section of the New Jersey Pine Barrens. According to Google Maps, there was a housing development in the middle of this vast stretch of untouched land, when in fact trees were the only inhabitants of this remote area. “Google Earth made a mistake, and I wanted to get in there,” says Vigg.

In Vigg’s words, his drones help him “describe contemporary landscape and the changes in current aesthetics while investigating the production of space.” In his most recent work, Vigg pushes the limits of what photography can do with an eye toward traditional landscapes. In order to capture the precise images that he wanted (when conventional means of capturing aerial shots – like kites and balloons – weren’t cutting it) he made drones using the same high-tech equipment scientists are using in unmanned aerial vehicles.

“I’m looking to get the look of the drone. I want to use what everyone has access to. It’s about accepting what this technology is going to give us and looking at the aesthetics and the pleasing nature of that,” he says. In an age where Google Maps and Google Earth have become our preliminary peep show of physical experience, Vigg’s taking the concept a step further, venturing into territory where no Street View car has gone before. 

How did you get the idea to work with drones?

I came across several air bases [on Google Maps and Google Earth] that had drones on them, so I started making drawings of the drones. I needed an idea for more advanced aerial shots and here I’d already been investigating these top secret air bases. I’d had them on my brain for weeks. So I started researching how to produce my own. 

How did you build and program the drones?

I build my own frames because they hold the electronics I like. I source parts from all over the place. The bulk of it is from 3D Robotics. The programming is all open source. You literally click points on the Google Map to tell this thing where to fly. I use GoPro cameras. When I first started, it was with a GoPro 2. I was having trouble with the image resolution, but closer to the tail end is when the GoPro 3 came out. It offers much better image quality and can tether the drones to my iPhone, offering more maneuverability in the air. It helped a lot; it was really a saving grace. It took off so much technology from the drone itself. 

Read the rest here

Read more…

3689531384?profile=original

How about an OSD module with Object tracking? Unfortunately there is no open source product so far.

Long Ago Chris introduced us with ADEPT 3000 (Specs Sheet)

(http://www.diydrones.com/profiles/blogs/ges-new-adept-3000-miniature)

a miniature video tracker for mini UAV's from a defense product manufacturer GE

Here is a video providing a closer look to product and apparently manufacturer has plans to introduce it in commercial market ( If i got it right )

Check its video here

http://www.engineeringtv.com/video/GE-Miniaturizes-Video-Tracking

3689531502?profile=original

Read more…
Developer

The CanberraUAV Outback Challenge team has put together a package that will allow anyone to try the image recognition system we developed for our successful 2012 OBC entry. We hope that by releasing this in an easier to use form that other OBC teams and anyone interested in image recognition in UAVs will be able to reproduce what we've done and build upon it for their own systems.

The fundamental problem that our system is meant to solve is providing real-time recognition of objects on the ground from a UAV, while requiring only a low bandwidth radio link between the aircraft and the ground station. A lot of UAVs are run using analog video links from the plane to the ground, but we think that is not a good way to do many search and rescue and environmental tasks. The problems with an analog video link are:

  • the quality of the video is often greatly degraded by being sent over the analog link, making it hard to spot small objects
  • the task of an operator staring at a video screen for hours is much too hard. The operator may notice something flash by, then has to decide whether to rewind (possibly missing something else in the search) or to skip the object
  • in many (most?) parts of the world analog video links may violate local radio licensing regulations

We think the solution is for the aircraft to do an initial pass of image recognition to find "interesting" objects on the ground, and then to show small (low bandwidth) thumbnails of those objects on the ground station, overlaid on a satellite map of the area. The operator can then select which of these thumbnails to look at more closely, bringing down a full high resolution image around that object from the plane over the telemetry link when needed. When using this method the operator gets a complete overview of the search area, and can quickly focus on areas of interest, using the human to decide which of the objects that the computer shows up are worth investigating. That was the strategy we successfully used in our 2012 OBC flight.

Up to now the image recognition code we developed for that competition was fairly deeply embedded in the MAVProxy module we used, making it hard for other teams or UAV researchers to run the algorithm over their own data to see how it performs. So to address that we have now put together a set of packages that can be installed on Linux, Windows or MacOS to try the system for yourself.

The new tool is called geosearch.py, and is a standalone UAV search tool that operates offline on a set of images, using either EXIF positioning data in the images, or a MAVLink telemetry log for geo-referencing. It runs the same code that we use in our real-time system embedded in the aircraft, but converted to run as a local GUI tool on a desktop computer. Note that the algorithms we use are tuned for speed. They are not state of the art computer vision algorithms, but they are fast, allowing them to run in real-time on a low power ARM Linux box embedded in an aircraft. We originally tried to use algorithms from the excellent OpenCV system, but found them too slow for our application. We do however use some OpenCV helper functions in our code to make the system a bit simpler.

Release of data from our OBC flight

As a companion to this tool we have released the images captured from our OBC 2012 flight for anyone to download, along with the telemetry logs and other files needed to reproduce the result. The images are kindly hosted by CSIRO in Australia, and are available from the CSIRO Outback Challenge website.

On that site you will find a set of 20 zip files containing the raw images from our flight. If you download and extract those zip files you will end up with a directory called 'raw' which contains 17042 pgm image files. Those files are 8 bit raw Bayer grid files of 1280x960, which are the raw images from our PtGrey Chameleon camera. We're hoping those files will be useful to anyone wanting a decent sized set of UAV images for research, or for training an algorithm for the next Outback Challenge.

(NOTE: as of the time of writing the CSIRO site has the images as 17k separate image files, those should be converted to the 20 zip files for easier download soon).

We also provide a tool called pgm_convert.py which will take those pgm raw images and convert them to other formats, most commonly to lossless PNG files.

How to try our system for yourself

If you watch the video and want to try our system on your own data, or on the data from our OBC flight, then you will need to install a set of python packages that provide the basic system tools we use in our code.

On Windows you need Python 2.7, plus a set of extra python packages. To make life easier for windows users we have collected all of the packages you need in this directory. Note that this includes both the packages needed to run our tools, and the packages needed to build our tools for yourself for when you want to modify the code.On Linux installing the basic python tools is a bit easier. For example, on an Ubuntu Linux system the following command should be enough:

  sudo apt-get install python-wxgtk2.8 python-matplotlib python-numpy python-pyexiv2 python-opencv python-httplib2 libjpeg-turbo-progs

If you are using MacOS, then you should install a suitable Python 2.7 package first, then use easy_install to add the required additional packages.

Once you have the basic tools installed, the next step is to install the 3 CanberraUAV python packages, pymavlink, MAVProxy and cuav. You can get those packages from the python packaging index:

That will allow you to download tar balls or windows installers for each package. If you are going to modify the code, then you will be better off downloading the source from our github repository.

To build the packages yourself (instead of using the pre-built packages), you would use the following commands in each of the 3 package directories. For Linux or MacOS, use this:

  python setup.py build install --user

for Windows use this:

  python setup.py build -c mingw32 install

Note that for building on WIndows you also need to install the MinGW package.

How it all works

To see how it all works, you can either read the code (just follow the above links to our code repository) or you can watch a presentation I gave in January 2012 which explains the algorithm to some extent (video and slides).

We'd also encourage you to ask questions on the CanberraUAV mailing list, or on the Outback Challenge discussion group.

How much bandwidth does it use?

The aim of this system is to allow for real-time UAV search with only a low bandwidth telemetry link to the aircraft. It is possible to run the system with a 64 kbit/s ISM band radio, such as the excellent RFD900 radio.

To make this possible we developed a new network transport protocol called block_xmit (mp4-video, ogg-video, slides, implementation), which allows us to intermix low bandwidth thumbnail images and python objects with normal MAVLink telemetry over a single radio link. The system also scales up to higher bandwidth links, so if you happen to have a high bandwidth ethernet bridge (such as a Ubiquity Bullet) in your aircraft, and you are close enough to the ground station to have a good link, then the system will transmit real-time images that gives the effect of a digital video link. That capability can be seen in the demonstration video above.

The block_xmit protocol also prioritizes images based on the image score from the recognition algorithm, so you can set quite a low threshold for what images to send, and if a high scoring object (such as Joe!) turns up it will jump the queue in the radio link and be shown to the operator quickly.

For our aircraft we combine both types of radio link, with the results of our imaging system coming over both a 5.8GHz Ubiquity ethernet bridge and a 900MHz RFD900 telemetry radio (with smaller thumbnail sizes to save bandwidth). That gives us redundancy in case a particular radio technology does not perform well on the day.

Geo-referencing

Included in the system is a geo-referencing system that uses the MAVLink telemetry stream along with image timestamps to work out the geographic location (latitude/longitude) of any region of interest in the image. These are displayed on the map in various formats (degrees decimal, degrees/minutes/seconds and UTM grid coordinates).

If you instead have a camera with a built-in GPS that tags images with EXIF geo-referencing tags the the code can instead use those for geo-referencing. That tends to result in much less accuracy however, as the EXIF tags don't contain nearly as much information about the aircraft attitude, which makes a big difference.

Also note that our current system uses local timestamps in the raw pgm filenames to work match the time to the MAVLink log. That is a mistake which we will fix in a future release. It means that if you are using our OBC data in a timezone other than Queensland you need to use the time offset option in the geosearch.py program to adjust the time to match the times in the MAVLink log. For example, if you are perth, which is 2 hours west of Queensland, then you would use an offset of -7200 seconds.

Next Steps

Some parts of our algorithm were tuned for the expected object sizes of the OBC competition (about 2 meters by 0.5 meters) and for the altitude and images sizes we used. We'd like to parametrize those aspects of the code to make it easier to use for a wider variety of tasks and camera systems.

We'd also like to make it easier to plug in different scoring filters. The recognition algorithm works in two stages, a first stage that finds anything unusual in the image, then a second stage that converts that region to HSV colour space and applies some simple heuristics to score the region. We'd like users to be able to easily plugin different scoring systems to suit their own image search task, as not everyone is looking for a lost hiker wearing blue jeans and a yellow shirt!

We are also interested in adapting our system to thermal cameras, such as the Tamarisk 640, and we hope to do some testing with that system later this year.

Please share your improvements

If you use this code in your own system then please share your improvements! You can clone our repository on github, or post an issue, or just discuss your changes on our mailing list. If we work together we can create a system which will end up saving lives, so please don't just keep your improvements to yourself.

Read more…
T3

3689531408?profile=originalThe above flight track highlights the performance of the latest version of the navigation algorithms in MatrixPilot. Flight tracks are from a flight that Peter Hollands conducted. (Thank you, Peter.) The red and the green tracks represent flight segments in opposite directions between the same pair of waypoints. The tracks are nearly perfect straight lines 243 meters long, with deviation between the two tracks of no more than 3 meters, in heavy winds. You will have to look closely at the picture, the two tracks are nearly on top of each other.

Winds are shown in the following picture.

3689531081?profile=originalThis precision navigation is the result of several recent innovations in the MatrixPilot algorithms. The main innovation is from Paul Bizard, and has to do with using IMU course over ground instead of heading for navigation. (Thank you Paul.) Paul has taken us full circle on this:

1. The first version of MatrixPilot used GPS course over ground for navigation. This had the advantage of not needing wind information, but suffered from GPS latency.

2. The next version of MatrixPilot used IMU heading for navigation. This eliminated GPS latency effects, but required wind to be included in navigation computations.

3. The next version of MatrixPilot used IMU for position estimation.

4. Recently, Paul pointed out that the MatrixPilot IMU furnishes course over ground, and that would be much better than heading for navigation. So we went full circle, we are back to course over ground. But now it comes from the IMU, not GPS, so there are no latency effects.

We also made several other improvements to the navigation algorithms:

1. We totally eliminated the use of angles in the navigation computations, everything is now based on vectors or matrices.

2. We are using 32 bit integer arithmetic in several places. This combines ultra-fine resolution with wide range.

3. We added a cross-track velocity damping term. This allows the cross tracking gains to be turned up without inducing a "dutch roll".

For more information on MatrixPilot, see its diydrones page, or its website.

Best regards,

Bill Premerlani

Read more…