A group of students in SRM University led by Sobin Santhosh,
final year student in the Aerospace department, has designed and implemented an
autonomous copter equipped with an automatic face and person recognition system
in a day. They have trained their intelligent software to detect a person
irrespective of his/her attire, face orientation, posture and/or background. In
their drill experiment of the rescue application, an onboard camera was mounted
on a quad-copter UAV where a live video was recorded. The video was fed to the
person identification software, developed on a MATLAB platform by Arijit Ray, a
final year student of SRM University in the Electrical and Electronics
Department. The software successfully detected the target person in an image
taken while the copter has hovering. Several images of the target individual
were taken in different postures, angles and backdrops to ensure that the
software has a robust accuracy. The GPS location of the snapshot image was also
transmitted to the ground station and it conformed exactly to the location at
which the snapshot was taken.Such an in-genuine application involving
autonomous copters enunciate the myriad scopes for further research and
development in the fields of smart security and surveillance. Rescue operations
that go on for days and even months can be completed in a matter of few hours
as copters can reach horizontal traversing speeds up to 5 m/s. Moreover,
autonomous copters can now be used as smart police surveillance aerial bots.
Latest research in image processing has made possible the detection of
activities in images. Such techniques can be further developed to be make more
robust under unrelenting conditions and then can be employed on copters for
smart detection of crime at weary hours of night when human surveillance is not
pragmatically possible.
Share to All
0 comments:
Post a Comment