Nom original: TestoftheraspberryPiinaplaneJuly12th2015.pdf
Ce document au format PDF 1.4 a été généré par , et a été envoyé sur fichier-pdf.fr le 14/07/2015 à 22:37, depuis l'adresse IP 198.128.x.x.
La présente page de téléchargement du fichier a été vue 338 fois.
Taille du document: 769 Ko (3 pages).
Confidentialité: fichier public
Aperçu du document
Test of the raspberry Pi in a plane : July 12th 2015
Goal: Take pictures of the ground from a high altitude with high speeds to see if we are able to
detect a fire.
Two raspberry Pi (the raspberry Pi called “Infra” and the other ), a power source (Sentey
Brio LS2152 ) two cable to connect the power source and the raspberry pi, two ethernet
cables and two laptops.
An airplane and its pilot
We have hooked the two Raspberry Pi and the power source on a black plastic square. We
have plugged in all the cables.
This system was stuck on the airplane’s belly thanks to a lot of tape. The ethernet cable went
through the door of the airplane. We plugged them in our computers inside the plane.
Description of the test:
We flew during approximately one hour above the east side of the bay (Berkeley, Oakland,
Richmond and surrounding areas) taking pictures of 8 targets, that is to say 8 people we had
started a fire in their backyard.
Each computer was connected to a camera controlling it thanks to the program “Picamera2.py”.
Unfortunately we lost the connexion with the nameless camera just after the takeoff. Hence we
only took pictures with the camera “infra” (with a blue filter which blocks the green light).
We took different kind of pictures. Indeed we ran the program several times with different
Pictures at 2240X1680 with 1.5 seconds delay between pictures.
Pictures at 640X480 with 1 second delay between pictures.
Thanks to this test we can obtain different informations. We can check the quality of our pictures
(Are they blurred ? Can we see the buildings on the ground ? ...). We can also check if we are
able to detect a fire.
We didn’t manage to take pictures of all the targets. However we could take several
pictures of Robin Lafever’s house where a fire was burning :
The fire is the very bright spot in the center of the image
We took a bunch of pictures of the ground with a high resolution (2240X1680):
This test answered to many questions. Indeed we now know that the images are not blurred
whatever the resolution. Moreover, we know that we are able to detect a fire from a high altitude
and a high speed.
Thanks to this test we are also now able to set our next goals. Indeed we need to improve our
system but also understand and fix what went wrong during the flight:
We need to understand why the “nameless” camera didn’t work and fix this problem
We need to replace the blue filter (which block the green light) on the camera “infra” by a
filter which only allow transmission above 730 nm to get infrared images.
We need to control both cameras with only one computer
We need to write a program to analyze the pictures. A program which can find the areas
of a picture where the pixels have particular properties (luminosity, colors,...), that is to
say areas areas where we could detect a fire.