High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Advanced Mechanism in Learning and Recognition of OPS from Weakly Labeled Street View Images

Author(s):

Vaishali A. Mahale , MET BKC,Adgaon,nashik; P. M. Yawalkar, MET BKC,Adgaon,Nashik

Keywords:

Real-World Objects, Street View Scenes, Learning and Recognition, Object Image Data Set

Abstract

Mobile phones are powerful image and video processing device containing the other various features like high-resolution cameras, colour displays, and hardware-accelerated graphics. The various applications give rise to a key technique of daily life visual object recognition. On premise sign is a popular form of commercial advertising, widely used in our day to day life. The OPSs containing visual diversity associate with complex environmental conditions. Observing that such, real-world characteristics are lacking in most of the existing image data sets. In this, first proposed an OPS data set, namely OPS-62, in which totally 4649 OPS images of 62 different businesses are collected from Google’s Street View. For addressing the problem of real-world OPS learning and recognition, developed a probabilistic framework based on the distributional clustering, in order to exploit the distributional information of each visual feature. Learning OPS images for more accurate recognitions and less false alarms. Experimental results shows that, using SURF descriptor technique applied on OPS-62 dataset outperform over the existing technique of using SIFT descriptor. Further, it shows marginal improvement in the average recognition rate. The proposed approach is simple, linear, and can be executed in a parallel fashion, making it practical and scalable for large-scale multimedia applications.

Other Details

Paper ID: IJSRDV4I50470
Published in: Volume : 4, Issue : 5
Publication Date: 01/08/2016
Page(s): 926-930

Article Preview

Download Article