Cattle in field

Plainsight Introduces Vision AI for Accurate Livestock Monitoring

The AI ​​platform can be used to both count the number of animals and monitor the health of the animals as needed.

Livestock in the field
Image: davidhewison/Adobe Stock

Animal husbandry is big business, and now Plainsight has: introduced an AI platform for helping farmers keep their livestock in the right numbers. The platform, known as Vision AI, has the ability to pinpoint the number of any given livestock, be it cattle, sheep or pigs in a particular area with 99.7% accuracy, according to Plainsight.

TO SEE: Ethical Policy for Artificial Intelligence (Tech Republic Premium)

“Customers came to us with this problem, and that’s typically how we build solutions,” said Elizabeth Spears, Plainsight’s chief product officer. “If you have situations where the livestock is not accurately counted, you can be susceptible to transport fraud within that process. There are a number of places in the supply chain where livestock counts can become inaccurate.”

Vision AI for livestock management

With livestock alone between $800 and $1,200 per head, it is imperative that those in the farming industry know exactly how much of any given animal they have on their property. While a large industrial farm was the inspiration for the artificial intelligence platform, Spears says this feature can be easily integrated into both large-scale and smaller family farms.

“I’d say it could work for both, but it’s most applicable on the larger farms, because that’s where counting usually becomes most economically essential,” Spears said. “It is a very practical solution. It is also very affordable. It’s most affordable for large scale operations, and the way it works is we usually work with the company to install it. So with Vision in particular, you have to maintain and update the model depending on your environment.”

To help combat this problem, Plainsight’s object counting system runs on a number of different components:

  • An accurate object detection model to classify and locate where the object is within each frame of a video. This usually requires training a custom model on a labeled data set.
  • An object tracking algorithm to track where each individual object is moving from frame to frame.
  • A registration area where object detection and tracking is applied when objects enter the camera image.
  • A counting line that triggers the object’s count as soon as they cross it.
  • An object moving backwards can be counted multiple times if you don’t take the direction of motion into account.
  • A deregistration zone where object detection and tracking can be removed with confidence after counting has taken place.

In addition to easily detecting the correct number of a type of livestock, the platform also helps farmers to ensure that animals remain healthy and safe on the farm.

“Counting is one thing it can do, but it can also really help monitor the health of the animal and also some sort of general surgery on the producer’s part,” Spears said. “You can look at the gait patterns of animals, so how they walk as an indicator of animal health, or food intake, and there are other kinds of visual indicators of health. It creates a way for producers to manage the larger numbers of livestock in general.”

Spears also says that if industrial or smaller farms want to install the Vision AI platform for their livestock management needs, the learning curve would be minimal.

“The platform shows very clearly what’s happening with your data, and that’s really essential to be able to track it over time,” she said. “One of the reasons people can have issues with the long-term accuracy of their models is that they don’t have a really clear understanding of the data coming in and the detections coming in. Accuracy is really essential, especially for this larger scale farms, because miscounting livestock is a tens of millions of dollars a year problem for them.”

Leave a Comment

Your email address will not be published. Required fields are marked *