Abstract: Accurate shadow elimination for video moving objects is an important issue in intelligent video monitoring，so a novel adaptive elimination algorithm is proposed to solve the problem.The optical characteristic of the shadow is analyzed firstly.The background reconstruction based on Gaussian mixture model and the shadow pre-detection are finished.Then，a novel adaptive dynamic texture analysis method is designed.Finally，the detection and elimination of the moving shadow is realized by the novel method.Experiment results prove the validity and practicability of the algorithm.

Key words: adaptive, dynamic texture analysis, moving objects, shadow elimination

摘要： 精确地消除活动阴影对运动目标的影响是智能视频监控的核心任务之一，对此提出了一种基于局部纹理分析的自适应阴影消除新算法。进行了基于高斯混合模型的背景重建，并根据阴影的光学特性进行了阴影区域的预检测，得到疑似阴影区域；提出了一种新的自适应动态纹理分析方法并在此基础上实现了活动阴影的检测与消除。实验结果验证了算法的有效性和实用性。

关键词: 自适应, 动态纹理分析, 运动目标, 阴影消除