We salute hackers who make technology useful for people in emerging markets. Leigh Johnson joined that select group when she accepted the challenge to build portable machine vision units that work offline and can be deployed for under $100 each. For hardware, a Raspberry Pi with camera plus screen can fit under that cost ceiling, and the software to give it sight is the focus of her 2018 Hackaday Superconference presentation. (Video also embedded below.)
The talk is a very concise 13 minutes, so Leigh flies through definitions of basic terms, before quickly naming TensorFlow and Keras as the tools she used. The time she saved here was spent on explaining what convolutional neural networks are and how they work, just enough to prepare the audience. But all of that is really just background, the meat of the talk is self-contained examples that Leigh has put together and made available online. I love to see that since it means you go beyond just watching and try it out for yourself. Continue reading “Leigh Johnson’s Guide To Machine Vision On Raspberry Pi”→
Some dogs have no sense of self-preservation. Given the opportunity, they will eat until they’re sick. It’s up to us humans to both feed them and remember doing it so they aren’t accidentally overfed. In a busy household with young children, the tricky part is the remembering.
Chloe’s kibble is kept in a touch-top wastebasket that flips open at the press of a button. [Bryan]’s dog-fed detector uses a reed switch and an Arduino clone to detect when the lid is opened. When the reed switch goes, low, the Arduino lights up an LED. The light stays on for two hours and then shuts off automatically to get ready for the next day. You don’t have to beg for a demo video, because it’s waiting for you after the break.
Since Chloe devours a bowl of food in about two minutes flat, maybe the next project for [Bryan]’s family could teach her to slow down a bit.
Dogs are remarkable creatures. Anybody who has lived with one will know that they are very vocal beasts, with barks that range from noting the presence of a squirrel in the yard to the warning whine that says “I am about to pee on your shoes if you don’t take me outside.” [Henry Conklin] decided to computerize the analysis of these noises, putting his dog [Oliver Twitch] on Twitter so he could hear what he was saying while he was at work. [Henry] that is: [Oliver] stays at home.
He did this using a Raspberry Pi, which is set to record sound above a certain volume. With the system sitting by [Oliver’s] favorite window, this records his barks. The recordings are then analyzed using PyAudioAnalysis, a library that analyzes sounds, compares them to reference ones and classifies them. The Raspberry Pi then posts the results onto twitter using Python-twitter.
Or rather, it will when [Henry] fixes a few bugs: right now it just posts a random string that is based on the length of the bark, not the type. [Henry] says he is working on the dog translation at the moment. It’s still a neat project that shows you how simple it is to use a few small bits of code to gather info from your environment and share these over the Internet. [Henry] also says that the next step is creating a weekly podcast for [Oliver]. I, for one, will be subscribing to hear his thoughts on how annoying the postman is, and how vexing it is to see a squirrel and not be able to chase them.
We’ve seen numerous products geared toward tracking the location and activities of your pets, two in the last month, but we feel sure you can make more functional devices than those you can purchase. Let’s look at a few and consider our options.