TSVD : Tunisia Street View Dataset

Today, the cameras are fixed everywhere, in streets, in vehicles, and in any public area. However, Analysis and extraction of information from images are required. Particularly, in autonomous vehicles and in smart applications that are developed to guide tourists. So, a large dataset of scene text images is an important and difficult factor in the extraction of textual information in natural images. It is the input to any computer vision system.

At present time, there is no available word-level annotated dataset of Arabic text  in natural scene images. Since our aim is to help the research community in standardizing the evaluation of scene Arabic text detection, the Tunisian Research Groups in Intelligent Machines of University of Sfax (REGIM lab of Sfax) will provide the Tunisia Street View Dataset  of Regim Lab (TSVD) freely of charge to mainly Arabic scene text detection researchers and to increase total of researches done to enhance scene text detection. The dataset contains 7k images collected from natural scenes.