Like me, I suspect that you have been, in turns, amused and puzzled by the recent outbreak of creepy clowns. But should you doubt the significance of this weirdness, realize that it’s not only a United States phenomena, it’s global! TIME just published an article titled The Creepy Clown Craze Is Now Taking Over the World while The Verge tried to be sort of rational with The 2016 clown panic: 10 questions asked and answered.
But no matter how much the media deny it, this is a serious thing, people. You thought the zombie apocalypse was a big deal? The clownpocalypse could well be Trump-sized, as in yuuuuuuggggggge. But fear not, dear reader, I have the answer as to how we can stay safe from marauding clowns; it’s a device called the Clowntector, an early warning system to spot and locate clown activity so professional, anti-clown professionals can do their sacred duty.
The Clowntector will be an Internet of Things (IoT) device based on the Raspberry Pi 3 Model B. It will work by grabbing images from a camera connected to the RPi then feeding them to a Python script that uses the OpenCV computer vision library to find faces in the images. Next, we’ll extract the faces, normalize their size, then hand them off to a neural network. If the neural network requires too much horsepower then we could offload the detected faces from the RPi and send them for processing on a more powerful machine. When a clown is detected, an alert will be sent out to a regional control center and the clown would be appropriately, er, "dealt with."
If you doubt whether this can be done, check out The fastest and bestest way to learn how to do computer image processing and A Practical Introduction to Deep Learning with Caffe and Python. The former, a book review I wrote last year of Practical Python and OpenCV + Case Studies which gives you a great insight into just how powerful computer vision can be on a Raspberry Pi (also check out the author's website for lots more examples and code), while the latter discusses using a convolutional neural network to distinguish between dogs and cats in photos; we'll use the same technology to distinguish between clowns and non-clowns.
At around $100 for the hardware (less in the huge volumes that would be required for effective coverage), the Clowntector would be cheap and could be deployed in public places such as malls, parks, and main streets as well as in schools, hospitals, and anywhere else evil clowns might appear to harass non-clown citizens. Just imagine thousands of Clowntectors configured to create a national Clown Detection Network (CDN). I’m thinking a Kickstarter or Indiegogo campaign to get the project rolling or maybe Homeland Security would fund the work.
With the Clown Detection Network in place, our children will no longer be frightened by large, red noses or old ladies terrified by the sight of a cream pie. My friends, safety from the creepy clown menace is within our grasp.