OpenCV in embedded systems: a performance insight
In “OpenCV in embedded systems: a performance insight,” Marco Di Vivo and Lorenzo Vangelista at the University of Padova in Italy and Michele Bevilacqua and Ivan Vecchiato from Embedia s.r.l.,describe an OpenCV-based algorithm running on a NEON enabled ARM-A8 processor based platform which has the task of finding an object in a picture, identifying its key visual features and looking it up in a database stored either locally or remotely on a server.
By using the OpenCV library, they were able to reduce the matching time needed to view the object in the picture and then accessing the database to identify it to a level that they think makes it acceptable for the real time usage of their application, which is to be ported for use on commercial mobile phone platforms.
Loading comments... Write a comment