CAJ | 학술논문

为了提高视频监控领域人数统计的准确性，提出一种基于团块分析的人数统计方法。首先通过光流算法获取前景团块的方向及能量强度信息，并结合团块大小等相关信息形成团块特征；然后针对人数统计提出一种新的目标跟踪算法；最后基于SVM对该团块特征进行训练分析，得到人数估计模型。实验结果表明，该方法正确率达到95％以上，能准确实现人数统计。
위료제고시빈감공영역인수통계적준학성，제출일충기우단괴분석적인수통계방법。수선통과광류산법획취전경단괴적방향급능량강도신식，병결합단괴대소등상관신식형성단괴특정；연후침대인수통계제출일충신적목표근종산법；최후기우SVM대해단괴특정진행훈련분석，득도인수고계모형。실험결과표명，해방법정학솔체도95％이상，능준학실현인수통계。
In order to improve the accuracy of people counting in video surveillance , the method for people counting based on the analysis of the mass was proposed .Firstly ,the direction and the energy intensity information of the foreground mass were obtained ,and combined with the size of foreground mass to form the feature of mass .Then ,the novel method of objects tracking was designed specially to aim at people counting .Finally ,the people counting model was obtained by training a support vector machine (SVM ) classifier with the input of the feature of mass .The experimental results showed that the accuracy of the counting was over 95% ,which confirms the method could accomplish people counting accurately .