Open AccessTSD CPI Traffic Sign Detection Technique Based on Centroid Position Identification in Text Mining
Author(s) -
R. Karthika,
S. Murugan
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.b3056.129219
Subject(s) - centroid , traffic sign , support vector machine , computer science , identification (biology) , cluster analysis , artificial intelligence , sign (mathematics) , pattern recognition (psychology) , position (finance) , distortion (music) , histogram , data mining , precision and recall , image (mathematics) , mathematics , mathematical analysis , amplifier , computer network , botany , finance , bandwidth (computing) , economics , biology
Detecting and Identifying traffic sign is a complicated issue due to the changing variability in cloud conditions. Hence, it is necessary to identify and detect of traffic signs during journey. The traffic text sign identification fails due to noise, blur, distortion and occlusion. In order to identify the text, a technique should be adapted that recognizes the text with improved accuracy. In existing algorithms such as Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) were not detecting the Centroid position. In this paper, the text Centroid of position sign is detected using text color, font and size. During journey, if the text is blurred, this Traffic Sign Detection Technique based on Centroid Position Identification (TSD-CPI) K-means algorithm for clustering is possible to use. As a result, it detects the text that with improved accuracy. Ultimately, it reduces the processing time. The experimental result reveals that using WEKA-3.8 with the proposed technique shows improvement over the existing algorithms in terms of precision and Recall which enhance the accuracy in text mining.