网球运动视频目标丢失点特征补偿方法研究
发布时间：2019-05-16 16:41
【摘要】：对于网球运动视频目标丢失点的特征补偿,传统方法有效性不足的缺陷,提出基于帧间差分消除的网球运动视频目标丢失点特征补偿方法。首先构建三维检测目标模型,利用帧间差分消除算法从运动视频图像中提取背景,预测目标运动轨迹,通过希尔伯特变换分析图像前景的轨迹相位差值特性,检测出目标丢失点。通过观察目标丢失点顶点结构,给出目标丢失点补偿措施。实验实测数据表明,所提方法对网球运动视频目标丢失点的分辨能力得到了大幅度提升,对丢失点的特征补偿精度更高,并且反应时间有效缩短。
[Abstract]:For the defect that the traditional method is not effective enough for the feature compensation of the lost point of the tennis video target, a feature compensation method of the lost point of the tennis sports video target based on the elimination of inter-frame difference is proposed. Firstly, the 3D detection target model is constructed, the background is extracted from the moving video image by using the inter-frame difference elimination algorithm, the moving trajectory of the target is predicted, and the phase difference characteristics of the foreground of the image are analyzed by Albert transform. The target loss point is detected. By observing the Vertex structure of the lost point of the target, the compensation measures of the lost point of the target are given. The experimental data show that the proposed method has greatly improved the resolution of the lost points of tennis video targets, the accuracy of feature compensation for lost points is higher, and the reaction time is effectively shortened.
【作者单位】： 河南工业大学;
【分类号】：TP391.41
本文编号：2478416
[Abstract]:For the defect that the traditional method is not effective enough for the feature compensation of the lost point of the tennis video target, a feature compensation method of the lost point of the tennis sports video target based on the elimination of inter-frame difference is proposed. Firstly, the 3D detection target model is constructed, the background is extracted from the moving video image by using the inter-frame difference elimination algorithm, the moving trajectory of the target is predicted, and the phase difference characteristics of the foreground of the image are analyzed by Albert transform. The target loss point is detected. By observing the Vertex structure of the lost point of the target, the compensation measures of the lost point of the target are given. The experimental data show that the proposed method has greatly improved the resolution of the lost points of tennis video targets, the accuracy of feature compensation for lost points is higher, and the reaction time is effectively shortened.
【作者单位】： 河南工业大学;
【分类号】：TP391.41
【相似文献】
相关硕士学位论文 前1条
1 海坤;智能监控中行人跟踪系统设计与实现[D];东南大学;2017年
，本文编号：2478416
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2478416.html
