Title: Classifying Ubiquitious data, images into Emotion for target Advertisement campaign
(get data from IP TV, youtube , and sensor network, correlate them with image processing data from image sensor for better ubiquity and targeted advertisement campaign
Objectives: Better Ubiquity and targeted advertisement by classifying images from CCTV other sensors at home like IPTV using image processing and correlating them with Advertisement campaign
Description of Project :
Advertisement campaign software today are basing there advertisement based on the cloud data at datacentres of gmail, YouTube etc.. But there is huge explosion of sensor data which was being generated by image sensors like CCTV on shopping malls, accepted webcam advertisement, gesture recognition, also the data from ubiquitous systems like home sensors internet of things
The data feeds from huge no of sensor at home as well the live IP TV were widening of eye pupil represent interest in particular advertisement on TV adv captured As TRP rating. Here people can get subsidised TV connection is they allow a webcam to record there sensor data which can be allowed for running Image processing algorithm on to classify different facts like emotions based on was reaction to adv running on TV or classification of interest. Which can be quantized and correlated with online data to find targeted advertisement campaign.
As Win win deal daily user can get feed of his emotion statistics throughout day will help him/her think about his responses throughout day and can be shown in behavioural software platform.
Also he and correlate and see trends of population in local area suppose limerick.
We are going to use cloud computing /hadoop to process image data and classify them using machine learning algorithms. And correlate with word interest feed from twitter or facebook to show trends.