(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Smart digital signage with eye tracking system / Chung Soon Zhi

Smart digital signage with eye tracking system_ Chung Soon Zhi_E3_2017_MFAR
Papan tanda digital adalah penyelesaian pengiklanan yang lebih berkesan berbanding dengan papan tanda tradisional kerana ia mampu untuk menunjukkan kandungan multimedia dan maklumat pengiklanan mudah untuk dikemaskini. Dengan kemajuan teknologi maklumat, sistem papan tanda pintar membolehkan beberapa interaksi antara penonton dan papan tanda. Dalam projek ini, satu sistem papan tanda digital pintar yang berkemampuan untuk berinteraksi dengan peranti mudah alih pengguna adalah dicadangkan. Aplikasi pengguna juga mengandungi fungi untuk mengemudi dan menyimpan maklumat iklan yang dipaparkan pada sistem papan tanda. Selain keupayaan interaktif, sistem yang dicadangkan juga dapat mengesan muka dan mata pengguna yang melihat papan tanda dan mengira tempoh penggunaan melihat setiap iklan yang ditunjukkan pada paparan. Pengesanan muka dan mata dibina dengan menggunakan pengelas “Haar Cascaded” yang terdapat di pustaka OpenCV. Satu algoritma “sanity check” dicadangkan untuk menapis pengesanan mata yang tersilap dan meningkatkan kadar pengesanan keseluruhannya. Keputusan ujian menunjukkan bahawa fungsi pengesanan muka dan mata yang dibina boleh mencapai 82.5% kadar positif benar dan 17.5% kadar negatif palsu. Ujian ke atas sistem yang lengkap menunjukkan bahawa sistem papan tanda pintar yang dicadangkan mampu berfungsi seperti yang dicadangkan. Ia mampu merekod tempoh tontonan pengguna untuk setiap iklan dengan ralat purata kurang daripada 12%. Digital signage is a more effective advertising solution compared to the traditional sign board since it is able to show multimedia contents and the advertising information can be easily updated. Current digital signage system has limited user interactive capability. Besides that, current system also lacks a way to collect viewer’s behaviour for analytic purposes. With the advancement of information technology, smart signage system allows some interactions between the viewer and the signage. In this project, a smart digital signage system which is capable of interaction between a user’s mobile device and the signage system is proposed. The user’s application on the mobile device provides a convenience way for navigating and storing the digital advertisements shown on the signage system instead of having paper brochure. Besides interactive capability, the proposed system is also able to detect faces and eyes to count the users viewing duration for each advertisement shown on the display. The faces and eyes detection were implemented using Haar Cascaded classifier available in the OpenCV library. A sanity checking algorithm is proposed to filter out the wrong detected eyes and improve the overall detection rate. Test results showed that the implemented face and eyes detection function can achieve 82.5% of true positive rate and 17.5% of false negative rate. Test on the complete system showed that the proposed smart signage system is able to work as expected. It is able to collect the users’ viewing duration for each advertisement with an average error of less than 12%.
Contributor(s):
Chung, Soon Zhi - Author
Primary Item Type:
Thesis
Language:
English
Subject Keywords:
Graphical user interface (GUI) ; interactive application (client).; haar cascade ; local binary pattern ; true positive rate (TPR), false positive rate (FPR) and false negative rate (FNR)
Sponsor - Description:
Pusat Pengajian Kejuruteraan Elektrik & Elektronik -
First presented to the public:
8/1/2017
Original Publication Date:
4/30/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 88
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-04-30 14:46:45.578
Date Last Updated
2020-05-29 18:28:35.566
Submitter:
Mohd Fadli Abd. Rahman

All Versions

Thumbnail Name Version Created Date
Smart digital signage with eye tracking system / Chung Soon Zhi1 2018-04-30 14:46:45.578