A California pet lover has come up with a way to adapt law enforcement's facial-recognition software to track down lost dogs.

John Polimeno got the idea when he saw a missing dog poster, triggering "horrible" memories of his lost dog. He developed the free Finding Rover app.

It works the same as the program used for people. Well, not exactly the same — it takes into account things like "length of ears" and "size of snout." But the theory is the same.

A dog owner downloads the app through a smartphone and registers their dog by snapping a picture. The app loads it into a database. If the dog goes missing, the dog owner clicks a button to report it. Meanwhile, animal-control agencies are snapping pictures of the stray dogs they've found. The software — using 128 facial features — searches the database for a match. If one comes up, the owner is notified.

You don't have to be a dogcatcher to track down the owner of a stray. If someone finds a dog, that person can take a picture of it and upload the photo to the site. If the dog is registered, the app will notify the person who found the dog.

Got a dog that doesn't like to pose for pictures? Polimeno thought of that, too. There's a "barking button" on the app that you can push to get the animal to look at you.

"It helps to get their attention," Polimeno said of the barking button.

Polimeno, 56, came up with the idea while at a coffee shop with his wife. He spotted a poster of a missing dog and it brought back memories of when his black Lab, Harley, went missing. He teamed up with the University of Utah. It took them about a year to develop the facial-features algorithm for pets.

By the way, for those who like a happy ending, Harley turned up again three days after he went AWOL. □