Cameras and AI can count, read and check things without recognising people. Useful examples for shops, cafés, clinics and clubs.
Computer vision means software that understands images. It gets attention for face recognition, which is the riskiest use. The more useful applications for a small business are much more ordinary, and most of them never need to know who anyone is.
Reading things
QR codes and barcodes are the simplest form of computer vision, and still one of the most useful. A scan can check in a booking, log a piece of equipment moving between rooms, or mark an order as picked up. Phone cameras do this well and cheaply.
Modern AI models also read text in photos: a receipt, a whiteboard of tomorrow’s schedule, a delivery note, a label on a box. That turns photos into records without retyping.
Checking things
- A café: a photo of the display case at closing to log what is left and plan tomorrow’s bake.
- A shop: checking shelf photos for gaps so restocking happens before customers notice.
- A clinic or studio: confirming that a room has been reset with the right equipment between bookings.
- A sports facility: photos of equipment condition at check-out and return.
Counting things
A sensor or camera that counts people coming through a door, without storing images or identifying anyone, can show your busiest hours and help with staffing. If you only need numbers, choose a system that only keeps numbers.
If the job is to count, read or check, the system rarely needs to know whose face it saw.
Good practice
- Point cameras at things, not people, wherever you can.
- Keep images only as long as the job needs, and prefer tools that process on the device.
- Put up clear signs where cameras are in use.
- Test accuracy in your own space: lighting, angles and clutter all matter.
Where a camera would capture people, privacy law applies, and recognition of individuals carries extra obligations. Our post on face recognition covers that in more detail.
From the DYOL blog. Photo by Yael Yañez on Unsplash.
Keep reading
All posts
Voice recognition has quietly got very good. Where it helps
Open speech-to-text models now run on modest hardware and handle many languages. Practical uses for phones, notes and orders, and what to watch for.

Face recognition at your business? Read this first
Facial recognition is cheap to buy and easy to switch on. In Canada it involves sensitive personal information, and the bar for using it is high.

What’s new in Claude Code: the late-summer releases that matter
Sessions that talk to each other, design in the terminal, background computer use and a way to test plugins. Here is what changed and why it matters.
A free 20-minute conversation, then an honest assessment. You’ll know what it costs and how long it takes, in writing, at a fixed price.
Or email lloyd.delacruz@outlook.com. We reply within 24 hours.
