A mechanism for learning and recognition of on-premise signs from street view images

Camera-enabled mobile devices are commonly used as interaction platforms for linking the user's virtual and physical worlds in commercial applications. The various application scenarios give rise to a key technique of daily life visual object recognition. On-premise signs (OPSs), a popular form of profitable advertising, are mostly used in our daily life. The OPSs often demonstrate great visual variety, accompanied with complex ecological conditions. In this paper we have used the OPS-62 dataset. The OPS-62 dataset contains 4649 OPS images of 62 different categories. We have used SIFT algorithm to extract feature like background, foreground, size etc. from OPS-62 images We have used distributional clustering for variable clustering. Experimental result shows the more accurate recognition.