随机加密图像中高效特定图像定位方法仿真

Simulation of efficiently Specific Image Positioning Method in Random Encryption Image

在随机加密图像中高效特定图像定位过程中,由于图像在加密后,源图像的像素变换会较为置乱,使得图像定位特征衰退明显,无法形成可以查询的规则。当前的图像定位算法,是通过提取图像共有特征的方法进行特定图像定位,加密后的图像不能精确剖析特定的细致特征,特征无法形成关联。为此,提出改进关联挖掘算法的随机加密图像中高效图像快速定位方法。对随机加密图像中所含有多余及重复的特定图像特征信息进行二次校验过滤,进而缩小定位特定图像错误范围,将在随机加密图像中定位特定图像的问题转换为关联规则挖掘中寻找频繁项集的问题,通过在随机加密图像中查找与特定图像的特征相似的频繁项集,完成对特定图像精准定位。仿真结果证明,提出的改进关联挖掘算法的高效图像快速定位方法精准度高,耗时短,为高效图像快速定位提供了科学手段。

An efficiently method for image＇s quick positioning in random encrypted images based on improved association mining algorithm is proposed. The specific image features information containing redundant and repetitive in random encrypted images are made second validation filtering,and the error scope of positioning specific image is shrunk. The question of positioning specific image in random encrypted images is transformed to the question of searching frequent item sets in association rules mining. Through the searching of the frequent item sets which are similar to the features of specific images in random encrypted images,the precision positioning of specific image is completed. The simulation result proves that the precision of improved efficient method for quick image positioning based on association mining algorithm is high,and the time-consuming is short.