高级检索

采用可逆信息隐藏技术的高性能彩色图像对比度增强算法

High Performance Color Image Contrast Enhancement Algorithm Using Reversible DataHiding

  • 摘要: : 针对彩色图像可逆信息隐藏中相对较少的嵌入容量和数据嵌入后的图像质量问题, 提出了基于可逆信息隐藏的高性能彩色图像对比度增强算法. 该算法将原始RGB图像转换成HSV(色调, 饱和度, 亮度)图像, 在预处理时合并像素直方图单侧2个高度最低的条柱并设置一个失真阈值防止图像失真; 数据嵌入过程在保持色调分量不变的同时增大了单个像素能嵌入秘密信息的长度, 且在增大数据嵌入量的同时减少了图像失真. 在多次执行数据嵌入过程后, 直方图得以实现均衡化, 从而有效增强图像的对比度. 在Kodak和USC-SICI图像集上的实验结果和分析表明,与同类算法(UCE和HSP算法)相比,所提算法在嵌入容量指标和色彩差异指标CIEDE2000的性能上分别平均提升了28.6%和34.7%.

     

    Abstract:  In order to solve the problem of less embedding quantity and image quality after a large amount of data embedding in reversible data hiding of color image, a high performance color image contrast enhancement algorithm based on reversible data hiding is proposed. In this algorithm, the original RGB image is converted into HSV(hue, saturation, value) image, in the preprocessing process, two bins with lowest height on the single side of pixel histogram are merged, and a distortion threshold is set to prevent image distortion; In the data embedding process, the hue component is maintained and the length of the secret information that a single pixel can embed is increased. Meanwhile, the image distortion is decreased. After several rounds of data embedding, the histogram is equalized, effectively enhancing the image contrast. Experimental results and analysis on Kodak and USC-SICI image sets show that compared with similar algorithms (UCE and HSP), the proposed algorithm improves the performance of embedding capacity and CIEDE2000 by 28.6% and 34.7% on average, respectively.

     

/

返回文章
返回