期刊文献+

儿童摔倒行为图像检测数学建模方法仿真 被引量:1

A Child Acts Fall Image Detection Method of Mathematical Modeling Simulation
下载PDF
导出
摘要 由于儿童对周围环境认知不健全,使得行为存在很多不确定性。传统的儿童摔倒行为图像检测过程中,主要通过采集儿童摔倒的前提行为特征进行建模分析,忽略了所处环境造成的不确定性对建模分析的影响,导致建模分析不准确的问题。为解决上述问题,提出基于仿射变换的儿童摔倒行为图像检测数学建模方法。通过背景减法获得儿童摔倒行为差值图像,结合Vally-Emphasis自适应阈值理论对差分图像进行二值化处理,并以灰度级作为分割图像的阈值,获得儿童摔倒行为差值图像通过自适应二值化后的前景图像,并组建六参数的仿射变换模型,利用上述模型估计出多帧儿童摔倒行为图像在下一时刻的匹配背景图像,将背景配准后的图像进行双向差分相乘,计算出儿童目标的质心位置,完成儿童摔倒行为图像检测。实验结果表明,所提方法能够有效提高儿童摔倒行为图像检测准确率,具有较好的鲁棒性和实时性。 The environment cognition of children is unsound. It makes the behavior have many uncertainties. In this paper, we proposed a mathematical modeling method for image detection of children tumble based on the affine transformation. Firstly, the author obtained the difference image of children tumble through background subtraction and conducted a binarization processing to difference image integrated with Vally-Emphasis self-adaption threshold value. The gray level was also used as the threshold value of segmented image to obtain the foreground image of children tumble difference image after self-adaption binarization processing. Then the affine transformation model with six parameters was built and the above model was used to estimate the matching background image of multi-frame children tumble image at next moment. Finally, two-way difference multiplication was made with the image after back- ground registration to work out the centroid position of children target and complete the image detection of children tumble. The experimental results show that the method mentioned above can improve the image detection precision of children tumble. It has better robustness and real-time property.
作者 王艳
机构地区 长春财经学院
出处 《计算机仿真》 CSCD 北大核心 2016年第11期452-455,共4页 Computer Simulation
关键词 儿童摔倒行为 图像检测 数学建模 Children tumble Image detection Mathematical modeling
  • 相关文献

参考文献10

二级参考文献114

共引文献106

同被引文献9

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部