In order to improve the recognition accuracy of sports video,and speed up the recognition speed of sports video, a sports video recognition model based on feature screening and support vector machine is proposed. The features of the sports video image are extracted,and input into the support vector machine for training. The average contribution value of each feature to the sports video recognition is obtained according to the training samples,and performed with descending sort to eliminate the useless and redundant features,and screen out the important features. The sports video recognition model is established according to the important features. The simulation experiment is used to test and analyze the validity of the model. The results show that the model has improved the recognition correctness of sports video,and its false recognition rate is lower than that of other contrast models.%为了提高体育视频识别的准确率,加快体育视频识别的速度,提出特征筛选和支持向量机的体育视频识别模型.首先提取体育视频图像的特征,并将特征输入到支持向量机进行训练;然后根据训练样本得到每一种特征对体育视频识别的平均贡献值,并根据平均贡献值进行降序排列,去除一些无用、冗余的特征,筛选出重要特征;最后根据重要特征建立体育视频识别模型,采用仿真实验对模型的有效性进行测试和分析.仿真结果表明,该模型提高了体育视频识别的正确率,而且误识率要低于对比模型.
展开▼
