path as well as wildlife cameras are commonly offered nowadays, however the Wild Eye project aims to go beyond just taking digital snapshots of critters. [Brenda Armour] utilizes a Raspberry Pi to not only take pictures of wildlife who roam into the camera’s field of view, however to likewise immediately determine as well as categorize the animals seen utilizing a visual recognition API from IBM via the Node-RED infrastructure. The result is a system that captures an picture when movement is detected, sends the picture to the visual recognition API, as well as attempts to determine any type of wildlife based on the returned data.
The visual recognition isn’t flawless, however a recent proof of idea shows appealing results with crows, a cat, as well as a dog having been effectively identified. perhaps when the job is prepared to move deeper into the woods, aspects from these solar-powered networked birdhouses (which likewise utilize the Raspberry Pi) could help cut some cords.
The HackadayPrize2017 is Sponsored by: