Date:23/04/18
The application uses a microphone built into the smartphone, an accelerometer and a gyroscope.
When the gadget collides with an object, the microphone detects an emerging sound unique to each material and the shape of the object, and the accelerometer and gyroscope record data on the displacement and rotation of the smartphone.
After completing the algorithm learning process, the creators of the application invited 15 volunteers who checked the accuracy of the program on 14 objects in quiet and noisy environments.
The participants were asked to tap the smartphone 50 times on each object. The analysis of the obtained data showed that the application operates with high accuracy. Thus, the worst rate of recognition was 94%, and the program identified some subjects with an accuracy of 99.7%. The results of the test also showed that external noise had little effect on the accuracy of identification.
In the future, the modified version of the program can find practical application in various fields. For example, by tapping a smartphone on a pan, the user will be able to start the countdown timer in order to remove the pan from the plate in time, or by tapping a smartphone on the handlebar of a bike the user will launch the navigation application on the smartphone, NEWSru reports.
New application allows you to identify an object by tapping it with a smartphone
A group of engineers from the Korean Institute of Advanced Technologies (KAIST) has created a mobile application that can identify objects by tapping a smartphone on their surface.The application uses a microphone built into the smartphone, an accelerometer and a gyroscope.
When the gadget collides with an object, the microphone detects an emerging sound unique to each material and the shape of the object, and the accelerometer and gyroscope record data on the displacement and rotation of the smartphone.
After completing the algorithm learning process, the creators of the application invited 15 volunteers who checked the accuracy of the program on 14 objects in quiet and noisy environments.
The participants were asked to tap the smartphone 50 times on each object. The analysis of the obtained data showed that the application operates with high accuracy. Thus, the worst rate of recognition was 94%, and the program identified some subjects with an accuracy of 99.7%. The results of the test also showed that external noise had little effect on the accuracy of identification.
In the future, the modified version of the program can find practical application in various fields. For example, by tapping a smartphone on a pan, the user will be able to start the countdown timer in order to remove the pan from the plate in time, or by tapping a smartphone on the handlebar of a bike the user will launch the navigation application on the smartphone, NEWSru reports.
Views: 1107
©ictnews.az. All rights reserved.Similar news
- The mobile sector continues its lead
- Facebook counted 600 million active users
- Cell phone testing laboratory is planned to be built in Azerbaijan
- Tablets and riders outfitted quickly with 3G/4G modems
- The number of digital TV channels will double to 24 units
- Tax proposal in China gets massive online feedback
- Malaysia to implement biometric system at all entry points
- Korea to build Green Technology Centre
- Cisco Poised to Help China Keep an Eye on Its Citizens
- 3G speed in Azerbaijan is higher than in UK
- Government of Canada Announces Investment in Green Innovation for Canada
- Electric cars in Azerbaijan
- Dominican Republic Govt Issues Cashless Benefits
- Spain raises €1.65bn from spectrum auction
- Camden Council boosts mobile security