计算机应用 ›› 2015, Vol. 35 ›› Issue (11): 3187-3193.DOI: 10.11772/j.issn.1001-9081.2015.11.3187

• 2015年全国开放式分布与并行计算学术年会(DPCS 2015)论文 • 上一篇    下一篇

基于图形处理器的视频二值概率分割

李金静1, 陈庆奎1,2, 刘宝平1, 刘伯成1,2   

  1. 1. 上海理工大学 光电信息与计算机工程学院, 上海 200093;
    2. 上海理工大学 管理学院, 上海 200093
  • 收稿日期:2015-06-17 修回日期:2015-07-20 发布日期:2015-11-13
  • 通讯作者: 陈庆奎(1966-),男,上海人,教授,博士生导师,博士,CCF高级会员,主要研究方向:网络计算、并行计算、物联网.
  • 作者简介:李金静(1988-),女,山东滨州人,硕士研究生,主要研究方向:并行计算、模式识别; 刘宝平(1990-),男,山东临沂人,硕士研究生,主要研究方向:并行计算、模式识别; 刘伯成(1982-),男,福建福州人,博士研究生,主要研究方向:高性能并行计算、CPU-GPU异构集群调度算法.
  • 基金资助:
    国家自然科学基金资助项目(60970012);高等学校博士学科点专项科研博导基金资助项目(20113120110008);上海重点科技攻关项目(14511107902);上海市工程中心建设项目(GCZX14014);上海智能家居大规模物联共性技术工程中心项目(GCZX14014);上海市一流学科建设项目(XTKX2012);沪江基金研究基地专项(C14001).

Binary probability segmentation of video based on graphics processing unit

LI Jinjing1, CHEN Qingkui1,2, LIU Baoping1, LIU Bocheng1,2   

  1. 1. College of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China;
    2. College of Management, University of Shanghai for Science and Technology, Shanghai 200093, China
  • Received:2015-06-17 Revised:2015-07-20 Published:2015-11-13

摘要: 针对现有视频二值分割算法分割性能过低的问题,提出了一种基于GPU的视频实时二值概率分割算法.该算法通过规范化视频帧中每个像素属于前景类和背景类的概率大小,实现了基于二次马尔可夫测量场(QMMF)模型的视频实时二值概率分割.首先分别为不同场景的视频帧提出了两种概率模型,即静态背景概率模型(SBLM)和动态背景概率模型(UBLM);然后,通过光照矫正算法颜色转换、阴影抑制算法阴影检测以及伪装检测算法来计算每个像素属于背景类的概率值;最后,通过Gauss-Seidel模型迭代计算出了使能量函数取得最小值的背景概率值进而得到像素的二值化值.此外,为了提高算法分割的准确性,该算法包含了对光照突变、投射阴影以及伪装情况的实时处理.同时,为了满足算法的实时性要求,在NVIDIA GPU上并行实现了该算法.验证了所提算法的分割性能即算法分割的正确性,测试了算法的GPU执行时间.实验结果表明,在算法分割完整性方面ViBe+和GMM+的平均漏检率和平均误检率分别是QMMF的3倍和6倍;在算法执行时间方面ViBe+和GMM+的平均GPU执行时间大约是QMMF的1.3倍.此外,还计算了QMMF算法的GPU加速比约为76.8.

关键词: 二值分割, 二次规划, 二次马尔可夫测量场模型, 概率, Gauss-Seidel模型, 统一计算设备架构

Abstract: Since the segmentation performance of existing binary segmentation algorithm for video is excessively low, a binary probability segmentation algorithm in real-time based on Graphics Processing Unit (GPU) was proposed. The algorithm implemented a probabilistic segmentation based on the Quadratic Markov Measure Field (QMMF) model by regularizing the likelihood of each pixel of frame belonging to forground class or background class. In this algorithm, first two kinds of likelihood models, Static Background Likelihood Model (SBLM) and Unstable Background Likelihood Model (UBLM) were proposed. Secondly, the probability of each pixel belonging to background was computed by tonal transforming, cast shadow detecting and camouflage detecting algorithm. Finally, the probability of background which makes the energy function have a minimum value was computed by Gauss-Seidel model iteration and the binary value of each pixel was calculated. Moreover, illumination change, cast shadow and camouflage were included to improve the accuracy of segmentation algorithm. In order to fulfill the real-time requirement, a parallel version of our algorithm was implemented in a NVIDIA GPU. The accuracy and GPU execution time of the segmentation algorithm were analyzed. The experimental results show that the average missing rate and false detection rate of ViBe+ and GMM+ are 3 and 6 times those of QMMF, the average execution time of GPU of ViBe+ and GMM+ is about 1.3 times that of QMMF. Moreover, the average speedup of algorithm was computed and it is about 76.8.

Key words: binary segmentation, Quadratic Programming (QP), Quadratic Markov Measure Field (QMMF) model, likelihood, Gauss-Seidel model, Compute Unified Device Architecture (CUDA)

中图分类号: