Computer Recognizes Humans By Their Walking Patterns

Posted: Aug 19 2005, 4:02am EST | by , Updated: Aug 10 2010, 4:22pm CDT, in News | Other Stuff

Computer Recognizes Humans by their Walking Patterns
/* Story Top Left 2010 300x250, created 7/15/10 */ google_ad_slot = "8340327155";

Don't Miss: This How to find Fingerlings in Stock

Finger prints, iris scans, facial features, vein pattern and now walking styles can be used to recognize individuals.

Nikkei.Net (Subscription) reports about a system developed by University of Tokyo's Center for Spatial Information Science (site) that can identify individuals based on their walk pattern in crowds.
The walk pattern recognizer combines floor-level laser beams with overhead cameras to isolate and track individuals.
Tests showed that monitoring 600m2 of a crowded train station, the system succeeded in isolating 80% of 150 poeple.
The system can be used to spot suspicious persons or measure performance of ads or signs in public spaces.
I do not think the system can identify a specific person by his or her walking style.See also the Omron Face Walker.

Holiday Gift Guides and Deals

Get your Holiday gifting inspired by Best Toy Gifts with High STEM Value and the Top 10 toy gifts under $10 if you are on budget. The most popular Holiday 2017 toy list include Fingerlings, Crate Creatures and more. Don't miss the new Holiday deals on Amazon Devices, including $29.99 Fire tablet.

This story may contain affiliate links.

This free App Solves You Holiday Shopping Problem

Download the free Tracker app now to get in-stock alerts on Fingerling, Luvabella, SNES Classic and more.

Latest News


The Author

<a href="/latest_stories/all/all/2" rel="author">Luigi Lugmayr</a>
Luigi Lugmayr () is the founding chief Editor of I4U News and brings over 15 years experience in the technology field to the ever evolving and exciting world of gadgets. He started I4U News back in 2000 and evolved it into vibrant technology magazine.
Luigi can be contacted directly at




comments powered by Disqus