Dogs have been said to be man’s best friend, but certain breeds somehow lack the directional sense so much so that they tend to get lost easily after bolting out from the house. Well, good thing there’s an app for that – in the form of Finding Rover, that is. Finding Rover would rely on a patent pending approach that they call Pet Match, where it will mathematically derive a metric for directly comparing dogs by their ID images. By taking the facial recognition approach that has been adapted from the active research fields of Machine Learning and Computer Vision, the Pet Match algorithm will be able to pick out key discriminating features among a midst of lost dogs. The algorithm would have the main say in these key features, and will most likely also contain other information such as the shape and position of the eyes, distinctive markings like stripes or spots, and the color of the fur.
Other than that, size, gender, breed, and location are additional information that will be used before facial recognition kicks in, in order to narrow the search parameters where lost dogs are concerned. Using Finding Rover is a snap – registration is done through Facebook and email, where you should have a decent photo of your pooch and a zip code.