【摘要】：對于網(wǎng)球運動視頻目標丟失點的特征補償,傳統(tǒng)方法有效性不足的缺陷,提出基于幀間差分消除的網(wǎng)球運動視頻目標丟失點特征補償方法。首先構建三維檢測目標模型,利用幀間差分消除算法從運動視頻圖像中提取背景,預測目標運動軌跡,通過希爾伯特變換分析圖像前景的軌跡相位差值特性,檢測出目標丟失點。通過觀察目標丟失點頂點結構,給出目標丟失點補償措施。實驗實測數(shù)據(jù)表明,所提方法對網(wǎng)球運動視頻目標丟失點的分辨能力得到了大幅度提升,對丟失點的特征補償精度更高,并且反應時間有效縮短。
[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.
【作者單位】： 河南工業(yè)大學;
【分類號】：TP391.41

【相似文獻】

相關碩士學位論文 前1條

1 海坤;智能監(jiān)控中行人跟蹤系統(tǒng)設計與實現(xiàn)[D];東南大學;2017年

，

本文編號：2478416