多特征检测耦合混沌映射的红外图像加密算法  

Infrared image encryption algorithm based on multi-feature detection and chaotic map

作　　者：彭英杰[1] 陈豪颉 PENG Ying-jie CHEN Hao - xie(College of Computer, Qinghai Nationalities University, Xining 810007, China College of Computer and Communication, Lanzhou University of Technology, Lanzhou 730050,China)

机构地区：[1]青海民族大学计算机学院,青海西宁810007 [2]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《计算机工程与设计》2017年第11期3099-3105,共7页Computer Engineering and Design

基　　金：青海省自然科学基金项目(2014-Z-618);青海省科技支撑基金项目(2015-ZJ-409)

摘　　要：为克服当前选择性加密技术易外泄密文目标的形状,导致其不能有效实现红外图像的安全传输的问题,设计基于多特征检测模型与低维复合映射的红外目标选择加密算法。引入形态学梯度,增大真实目标与背景的对比度差异;考虑红外目标与背景的灰度差异,改进Top-Hat变换,对其完成检测;构建多特征检测模型,获取包含红外目标的感兴趣区域;将一维Logistic映射作为触发器,联合sine映射、Tent映射,设计复合映射,改变感兴趣区域内的像素位置;改变复合映射的初值,输出新的混沌数组,设计加密函数,输出扩散密文。测试结果表明,与当前选择性加密技术相比,该算法能够更好地用于红外目标的加密,且其安全性更高。To solve these defects such as leaking the target shape of cipher and difficult to be used for infrared target encryption of the current image selective encryption algorithm,the infrared image selective encryption algorithm based on multi-feature detection model and low dimensional compound chaotic map was proposed.The contrast between the real target and the background was increased using morphological gradient.The Top-Hat transform was improved by considering the difference between the infrared target and the background for separating the infrared target from the background.The multi-feature detection model was constructed for obtaining the region of interest of the infrared target.The complex mapping was designed by taking one-dimensional Logistic mapping as trigger and combining with the sine mapping,Tent mapping for outputting random sequence.The encryption function was designed based on the new random sequence induced using the new initial condition for outputting the diffusion cipher.Test results show that the proposed algorithm can be better used for the encryption of infrared target with higher security and key sensitivity compared with the current selective encryption technology.

关 键 词：红外图像 选择性加密 低维复合映射 TOP-HAT变换 多特征检测模型 感兴趣区域 形态学梯度 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]