Systems and methods for automatically generating scoring scenarios with video of event
摘要
說明書
The object recognition module may use edge detection, pattern recognition, including, but not limited to, self-learning systems (e.g., neural networks), optical character recognition, on-line character recognition (including but not limited to, dynamic character recognition, real-time character recognition, intelligent character recognition), and/or any other suitable technique or method to determine the objects and/or characteristics in media assets. For example, the media application may receive sports programming in the form of a video. The video may include a series of frames. For each frame of the video, the media application may use an object recognition module to determine the characteristics associated with each player currently appearing in the sports programming.
In addition, the media application may use multiple types of optical character recognition and/or fuzzy logic, for example, when comparing multiple data fields (e.g., as contained in data structures described below). For example, the particular data field may be a textual data field. Using fuzzy logic, the system may determine two fields and/or values to be identical even though the substance of the data field or value (e.g., two different spellings) is not identical. In some embodiments, the system may analyze particular data fields of a data structure or media asset frame for particular values or text. The data fields could be associated with characteristics, additional information, and/or any other data required for the function of the embodiments described herein. Furthermore, the data fields could contain values (e.g., the data fields could be expressed in binary or any other suitable code or programming language).
