Text detection in video frames using hybrid features

Text in video frames provides brief and important content information which is helpful to video scene understanding, annotation and searching. A new text detection method in video frames is proposed in this paper. First, a small overlapped sliding window is scanned over the frame from which hybrid features are extracted. And then SVM classifier is employed to distinguish the text from background. At last, vote mechanism and morphological filter are performed to precisely locate the text region. The new method is expected to outperform the existing strategies based on the following two improvements. One is selecting robust features to distinguish both the scene and overlay text from the complex backgrounds. The other is addressing the multilingual capability over the whole processing. The proposed algorithm has been evaluated by four different kinds of videos and the experiments show its high performance.