Detection of Bank Logos on Video using Faster R-CNN Method

Abstract

In recent years, deep learning methods have achieved high success as solution to problems in the computer vision. Especially, CNN algorithms that extract information from the image is widely applied in logo detection. In this case, the recognition of trademark in trademark applications or infringement has been one of the major problems in the literature in terms of companies. In this paper, the dataset containing the logos of banks acquired from public domain images was collected in order to perform logo recognition, by using Faster R-CNN, an approach for the recognition of the bank logo in the video have been developed and as a result average accuracy of %98 was obtained.