Comments

• 

  Vega Tech Group LLC December 10, 2014 at 5:59am

  The issue is the hardware itself. Poor quality digital imaging sensors create these gaps, miscues, and poor data results. With the proper imaging sensor, lenses, camera electronics designed specifically for the application there is no need to correct for clouds, light breaks, even bad weather.

  This is the real issue with the current hardware being used. Consumer grade imaging sensors, cameras, and lenses are not made for quality data, they are made to take pretty pictures.

  No matter what camera being used, Canon, Sony, Nikon, etc that has had its Visible filter removed and replaced with something else, you still have a consumer grade sensor, lens, electronics. 
• 

  Jeff Foster December 9, 2014 at 7:05pm

  Wow - great article RIchard Burd! Thanks for sharing!
• 

  Vega Tech Group LLC November 1, 2014 at 8:27pm

  Realtime image analysis is currently in use, has been for some time. We support several cameras with embedded computers, both MS & Linux. Image processing is onboard, using wide bandwidth wireless, realtime processed images cuts flight time as well.

  Instead of multiple flights to build a data base, data can be achieved in a single pass.
• 

  armahdi hatta November 1, 2014 at 8:51am

  UAV ,, U CAN SELL CONTROL FPV 10 KM
• Leonard Maree May 24, 2014 at 9:37am

  Hi .. Name Leo Currently Sanae Base Antarctica. Transforming Electric Glider to Drone. Saw same requirements in NAM few months ago. Like to get involved.
• 

  james sowell May 24, 2014 at 8:17am

  http://www.agweb.com/drone_zone.aspx 
• 

  Deon van der Merwe December 16, 2013 at 5:03am

  To Milan, we are actively doing research on toxic algae blooms at Kansas State University, including the determination of algae density using small unmanned systems. This video demonstrates such a flight:
  http://youtu.be/Hu28ltyBxro
  
  The video was shot at a site where we have a COA.
  The coverage is limited by the duration of the aircraft and the typical regulatory restrictions of flying line of sight.
• 

  Michael Leasure December 16, 2013 at 3:15am

  Milan, the algae was NVDI using a Sony cybershot with custom filter and lense. Really big and heavy cam, old tech, but cheap. That was done about 7 years ago. The study was of the rain runoff and its impact on the bloom. I don't understand your question on the 40 acres. The camera captures that much acreage per photo and they are patched together back at the office. The critical issue was that the photos be straight down and not angled at all. This required piezo gyro stabilization back in the day. Now the ardupilot can take care of that. In actual field deployment, simple equated to flexible and reliable. I know this is bad news for the robotics, autonomous types but we found a GPS downlink to a laptop, stabilization gyros, and a manually triggered camera worked most reliably. More autonomy than that and we were fighting the glitches, wind, radio links, etc. i am not yet convinced autopilots and automated camera track and triggering are needed or reliable. We continue to test and modify but often uncover weaknesses in field use. Keep it simple and build on that.
• MILAN BEHARIE December 16, 2013 at 2:14am

  Hi Micheal, I would like to become more involved in this field it interests me. Regarding algae blooms do you use NVDI? Also is the monitoring of these purely for the benchmarking of treatments? Also 40 acres is 0.2km^2 and this entails whats kind of flight path? Is it perimeter or sectional?
• 

  Michael Leasure December 15, 2013 at 2:42pm

  To Milan below, we have flown algae blooms, 3D model generation of micro topography using images, marijuana detection, stress in mint crops, drought damage, underground tile location, effects on bio mass with and without fungicide application for Monsanto, effects of weed shading on new plant growth, and several generic health evaluations of corn and soybean crops. We do 40 acres at a time, IR and color. Purdue University is a leader in image processing, regardless of the source of the image.