Skip to Content
Uncategorized

Smart Cameras

A new tool can decipher text in video.

Government and private surveillance companies have a new weapon. Software developed at SRI International in Menlo Park, CA, can identify words and numbers in moving video taken under difficult conditions. The software locates characters, extracts them from the background, and adjusts for lighting or angled views. Gregory Myers, program director at SRI, says the software could be used in applications ranging from video archiving to homeland security.

The software works by examining each frame of a video for sharp lines set against a contrasting background – telltale indicators of character strokes. Since text often persists over many frames, the software also looks for text similarity among frames to increase accuracy. If characters are warped or unclear, image-processing algorithms straighten them and even out their tone. The cleaned-up characters are then fed into optical character-recognition engines that convert them to machine-searchable text.

The company is working with government agencies that need to identify ships and vehicles from rough video; Myers expects that the software will soon be available for private consumer use as well.

Keep Reading

Most Popular

conceptual illustration showing various women's faces being scanned
conceptual illustration showing various women's faces being scanned

A horrifying new AI app swaps women into porn videos with a click

Deepfake researchers have long feared the day this would arrive.

A view of clouds illuminated by sunlight
A view of clouds illuminated by sunlight

We can’t afford to stop solar geoengineering research

It is the wrong time to take this strategy for combating climate change off the table.

Death and Jeff Bezos
Death and Jeff Bezos

Meet Altos Labs, Silicon Valley’s latest wild bet on living forever

Funders of a deep-pocketed new "rejuvenation" startup are said to include Jeff Bezos and Yuri Milner.

new GPT3 is a good student
new GPT3 is a good student

The new version of GPT-3 is much better behaved (and should be less toxic)

OpenAI has trained its flagship language model to follow instructions, making it spit out less unwanted text—but there's still a way to go.

Stay connected

Illustration by Rose WongIllustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.