面向移动终端的视频图像模糊特征提取仿真

摘要

对移动终端的视频图像的模糊特征进行提取,能够有效改善移动终端视频图像低照度的图像质量.对移动终端的视频图像特征进行提取,需要获得图像特征初匹配点对,进而对特征单向匹配结果进行双向匹配.传统方法构造移动终端视频图像模糊特征指标,将视频图像细节信息进行划分,但忽略了特征点的双向匹配,导致提取精度偏低.提出基于SURF的虚拟现实技术下移动终端视频图像模糊特征提取方法.对输入的移动终端视频图像进行视频图像灰度提取,检测移动终端视频图像中的闭合区域作为终端视频图像的放射不变区域,获得移动终端视频图像初匹配点对,再对获得的视频图像模糊特征单向匹配结果进行双向匹配,最后对误匹配点进行剔除,由此完成移动终端视频图像模糊特征提取.实验结果表明,所提方法能够有效改善移动终端视频图像的视觉效果,提升亮度、清晰度.%An extraction method for image fuzzy feature of mobile terminal video under the virtual reality technology is proposed based on SURF.Firstly,the gray scale of video image for video image of input mobile terminal is extracted and closed region in a video image of mobile terminal is detected as radiation invariant region of terminal video image to obtain initial matching point of video image of mobile terminal,and then bidirectional matching for the results of one-way matching of fuzzy features of video images is carried out.Finally,the mismatching points are eliminated and fuzzy feature extraction of video image of mobile terminal is completed.The simulation results demonstrate that the proposed method can effectively improve visual effect of video image of mobile terminal and promote brightness and clarity.