陈培培, 顾韵华. Android平台下基于ROI运动检测的视频字符实时定位[J]. 微电子学与计算机, 2013, 30(8): 86-89,93.
引用本文: 陈培培, 顾韵华. Android平台下基于ROI运动检测的视频字符实时定位[J]. 微电子学与计算机, 2013, 30(8): 86-89,93.
CHEN Pei-pei, GU Yun-hua. Real-Time Video Character Locating Algorithm Based on ROI Motion Detection on Android Platform[J]. Microelectronics & Computer, 2013, 30(8): 86-89,93.
Citation: CHEN Pei-pei, GU Yun-hua. Real-Time Video Character Locating Algorithm Based on ROI Motion Detection on Android Platform[J]. Microelectronics & Computer, 2013, 30(8): 86-89,93.

Android平台下基于ROI运动检测的视频字符实时定位

Real-Time Video Character Locating Algorithm Based on ROI Motion Detection on Android Platform

  • 摘要: 提出了一种 Android 平台简单背景下基于兴趣区域(Region of Interesting,ROI)运动检测的视频字符实时定位方法。该方法考虑到视频图像的连续性和相似性特点,采用区域运动检测跟踪设备位置,由初始帧定位获得感兴趣区域,然后对相邻帧计算 ROI 状态变化⊿,仅对⊿大于阈值的帧执行字符定位过程,其余帧则沿用前帧定位结果。而帧字符的定位主要采用数学形态学与连通区域相结合的方法进行。实验结果表明,提出的方法有效地提高了移动平台下字符定位的速度。

     

    Abstract: In this paper, an algorithm of real-time video character locating based on ROI motion detection is presented to solve the low efficiency problem of character locating in video shot by mobile devices.Considering the video images' features of discontinuity and similarity, this method uses regional motion detection to track the position of equipment and get the region of interesting (ROI) from the result of first frame character detection,then calculate the changes of ROI state (marked with ⊿) of adjacent frames.Only when ⊿ is greater than the threshold value,the method relocates the frame.The others will use the location result of last frame.The character locating algorithm mainly uses the combination of mathematical morphology and connected region analysis.The experimental results indicate that the proposed method can effectively improve character locating speed in mobile platform.

     

/

返回文章
返回