A system for automatic text detection in video

Text information extraction in images and video: a survey

“…They achieved an accuracy of 90% in scene-change detection. Gargi et al [29,32] performed text detection using the assumption that the number of intracoded blocks in P-and B-frames of an MPEG compressed video increases, when a text caption appears. Lim et al [33] made a simple assumption that text usually has a higher intensity than the background.…”

Text detection, localization, and tracking in compressed video

“…Unlike the text tracking methods in Refs. [10,19] which are carried out in the pixel domain, the motion vector information is utilized to track the detected text lines in compressed domain according to the motion similarity of the MBs in the text regions [36].…”

“…Text tracking stage can be used to verify the text localization results [10,19,36]. The appearing and disappearing frames of each text line in a video sequence will provide important clues for highlight events detection in semantic-based video analysis, indexing and retrieval [3][4][5][6].…”

A spatial-temporal approach for video caption detection and recognition

“…To construct such systems, both low-level features such as object shape, region intensity, color, texture, motion descriptors, audio measurements, and high-level techniques such as human face detection, speaker identification, and character recognition have been studied for indexing and retrieving image and video information in recent years [3], [4], [10], [11], [13], [19], [21], [24], [27]- [29], [32], [36]. Among these techniques, video caption based methods have attracted particular attention due to the rich content information contained in caption text [1], [2], [6], [9], [11]- [13], [15], [16], [19], [20], [27], [33], [36]. Caption text routinely provides such valuable indexing information as scene locations, speaker names, program introductions, sports scores, special announcements, dates and time.…”