Xyris 6 with its thermographic payload in cattle tracking project
There is a huge number of UAV projects and another new appear basically every day. Some of them are daring, groundbreaking or even utopian, another are more likely a solid utilization of already existing technology, method and equipment within new conditions. In this blog, we have already discussed quite a lot topics about precision agriculture and aerial thermography. Moreover, we are also focused on providing equipment and know-how for this UAV applications. However, since precision agriculture with drones usually means to scan and measure conditions of crops, there are also farmers who are tracking cattle with aerial thermal imaging. Or at least they could do it. This is actually a goal of project of two Canadian students, Chris Foster and Matt McInnes. We should also mention that they used our hexacopter, Xyris 6, as a one of two UAV configurations for the project.
These two students have designed their project with a support of two supervisors, Kevin O’Neil and John Church. Their starting point statement was that finding the cattle at very spacious ranchlad is usually difficult and it requires teams of people and significant time. In addition, technology has yet to do much in this field. In summary, cattle roundup is just expensive and time consuming and tracking cattle with aerial thermal imaging can be a great solution to this problem. This UAV application has two main parts: a multicopter, equipped with fully autonomous navigation and controlled by a ground station and an optic payload. In this very case the team had choosen Xyris 6 and BigX (quadcopter), including PixHawk Flight Controller and GPS. Tracking cattle with thermal imaging was ensured with FLIR A310 IR camera and mounted comptuter. The team also used two softwares for image processing and analysis, Python and OpenCV.
The actual algorithm was pretty simple. Firstly, an image from the video source has to be pulled. Secondly, you should blur the image slightly. Then you have to filter everything below treshold temperature. Then define countour region. If contour regions exist, query the GPS location Output/log the GPS location and repeat. According to the team´s project report, they met some pretty challenging tasks during the research, but they learned a lot about hardware and expanded their breadth of knowledge. They also realized that flying the drone isn´t always that easy as they think, but it is manageable in time. Thanks to the gathered experience they were able to suggest two main options, regarding the commercial potential of tracking cattle with aerial thermal imaging: farmers can either fly their own drones, which is not so easy for everyone, or they can purchase services of trained and certified pilots. Moreover, they are also decided to take another steps in their project in future, such as implement software on Xyris 6, improve detection accuracy, implement real-time tracking and test it on more cows. I´m looking forward to hopefully read some good news from their research soon.
If you are interested in this promising project, you can read more informations about tracking cattle with infrared imaging right here.