【作者】 杨雪； 于晓洋； 王霞； 张健； 杨秀坤；

【Author】 Yang Xue~1 Yu Xiaoyang~1 Wang Xia~1 Zhang Jian~2 Yang Xiukun~3 (1 College of Measurement Control Tech & Communication Engineering Harbin University of Science and Technology Harbin 150040 China;2 College of Information & Computer Engineering Northeast Forestry University Harbin 150000 China; 3 MOTOROLA countries of Southern California to identify R &D Center USA)

【机构】 哈尔滨理工大学 测控技术与通信工程学院； 东北林业大学信息与计算机工程学院； 美国南加州MOTOROLA国家识别研发中心；

【摘要】 鉴于传统遗传算法用于块截断编码图像信息隐藏时存在早熟和稳定性差的缺点而使得原始彩色图像与重构图像均方误差大,提出了用于块截断编码图像信息隐藏的遗传算法(BTCIIHGA)。首先对块截断编码产生的二进制位图种群初始化、均匀排序选择、自适应交叉、自适应变异,然后利用进化逆转操进一步优化种群得到通用位图。BTCIIHAGA算法产生通用位图来代替所有原始三个彩色位图,从而降低比特率、加快传输速度。实验结果表明,BTCIIHAGA算法比传统遗传算法的均方误差小、图像压缩率高、收敛速度快、稳定性高。更多还原

【Abstract】 Because of the standard GA have such disadvantages as premature convergence,low convergence speed and low robustness which generate the large general mean-square error between the original images and the common bitmaps.A genetic algorithm applying to block truncation coding image information hiding is proposed. The algorithm first initializes the binary bitmaps population generated by block truncation coding and then uniform ranking selection,adaptive crossover,adaptive mutation and reverse to optimize the population to get the common bitmap.At last,the results of MATLAB stimulations are given.BTCIlHGA generates the common bitmap to replace the original three bitmaps to reduce the bit rate and speed up the transmission.The experimental results demonstrate that the proposed method is practical for giving the minimum error and increasing the compress rate and the faster convergence speed and higher robustness.更多还原