Ad Selection on Videotron Using Yolo TensorFlow Based on Type of Vehicle Mochammad Arie Aldiansyah (a*), Yaddarabullah (a), Budi Arifitama (a)
a) Program Studi Teknik Informatika, Universitas Trilogi
Jl. TMP Kalibata No.1, Jakarta 12760, Indonesia
* mariealdiansyah[at]trilogi.ac.id
Abstract
Outdoor media advertising has a business potential of up to IDR 66.8 trillion per year, even urban society has a level of consumption in seeing outdoor promotion reaches 81%. One of the outdoor media advertisements that used by businesses is Videotron. However, Videotrons in ad serving still rotate sequentially and not appropriate based on the target market and the condition of the road at the location of the Videotron. To increase the effectiveness of Videotron to fit according to the target advertisement market based on its broadcasts can be done by creating a new system modelling that can determine adverts based on the target market and the conditions of the roads. This study will use Yolo TensorFlow which can detect the type of vehicle in one image and then supported by the Technique for Order of Preference by Similarity to Ideal Solution for determining the advertisements that will be displayed based on the condition of the roads. The results of this study obtained the accuracy of Yolo TensorFlow in detecting vehicles for the morning at 68.53%, then for the daytime at 83.17% while in the afternoon it was 30.20%, where to detect public vehicles has an average percentage accuracy of 100% while the average value the smallest mean is on a motorcycle that is 44.87%.
