经纬仪测量图像中目标缺损区域的恢复  

Restoration of Target Incomplete Area in Theodolite Measuring Image

作　　者：李晓冰[1] 高冰[1] 

机构地区：[1]92941部队,辽宁葫芦岛125000

出　　处：《飞行器测控学报》2007年第3期30-33,共4页Journal of Spacecraft TT&C Technology

摘　　要：根据电子经纬仪序列测量图像的特性，充分利用缺损帧前后数帧图像的相关信息，采用C-均值模糊聚类技术，提出了一种测量图像缺损部分恢复方法，对恢复帧前后数帧进行聚类，并剔除噪声帧，采用有效聚类中心灰度值对缺损区域进行恢复。经试验证明：此方法恢复的目标区域图像准确，灰度变化平缓，不失真。According to the characteristics of cine-theodolite serial measuring image, the paper effectually utilizes related information of incomplete image frame neighborhood by using fuzzy C-means clustering. A new method for restoration of measuring image incomplete area is presented. The image of restoration frame neighborhood is clustered, the noise frame is eliminated from the image, and the gray value of effective cluster center is used to restore incomplete area. The experimental result shows that the restoring target area is accurate by use of this method, the change of gray is gentle and without distortion.

关 键 词：序列图像 缺损图像 C-均值模糊聚类 灰度值 图像恢复 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] V557.4[自动化与计算机技术—计算机科学与技术]