Mapping sports interest with social network

Discovering regions that have sports interest in a set of images acquired from a scene at different times and possibly from different viewpoints and cameras is a crucial step for many applications. Physical activity can be effective at all stages of chronic disease, therefore, finding regions with the presence of physical activities might contribute to is important for the elaboration of public policies to minimize the presence of diseases such as obesity. This work addresses the problem of sport/non-sport image classification. We combine Convolutional Neural Network (CNN), traditional classifiers and geographical information to provide robust training and testing stages. As result, we achieved a high area under the curve (AUC) in a social network dataset. The experimental results show the feasibility of our proposed model. These results can be used and applied to develop public health policies based on statistics of sports interest.