Abstract: Wonderful event detection for soccer video has been the hot and difficult problem in the field of sport video semantic analysis.Based on the powerful expression ability of hidden conditional random fields (HCRF) model in the expression and identification of semantic event, this paper presents a novel framework for corner, penalty, and the red and yellow cards wonderful events detection based on the multidimensional semantic clues and HCRF model.Firstly, through analysis of the structural semantics of the wonderful event video, ten kinds of multidimensional semantic clues are defined to accurately describe the included semantic information of the wonderful events.Secondly, the video clips are split into several physical lens, then the multimodal semantic clues are extracted from the key frame of each lens to get the feature vector of the current lens, and the observed sequence is composed of the feature vectors of all shots in the test video clips.Finally, the above observed sequence is used as HCRF model input in the case of small-scale training samples, and a wonderful event detection HCRF model is effectively established.The inherent laws of the wonderful events are excavated from the multiple dimensions of the video structural semantics based on the mapping relationship of audio and video low-level features, the multidimensional semantic clues and the wonderful events, and the wonderful events;and the wonderful events detection is precisely achieved.Experiments show the effectiveness of this framework.