摘要
设计一种基于离散余弦变换的增嵌入式图像编码算法,首先对原始图像进行离散余弦变换,然后按特定扫描顺序对变换域余弦系数进行扫描以便能够很好地确定各系数的位置,其次为了用较少的编码位数表示变换域余弦系数需要对扫描后的系数进行量化。实验结果表明该图像编码算法可以根据目标码率或失真度大小的要求随时结束编码,并能解码近似恢复原图像;且随着压缩比的减小,即传输数据量的增加,无论是主观质量还是客观指标RMSE与PSNR都有较大改善。
This paper designs an embedded image coding algorithm based on discrete cosine transform.The coding process is to transform the image with discrete cosine transform firstly,and then to scan the cosine coefficient in the transform domain in a specific scanning order so that the positions of each coefficient can be well determined.Secondly,in order to represent the cosine coefficient in the transform domain with fewer coding bits,the scanned coefficient should be quantified.The algorithm can end the coding at any time according to the requirements of the target bits or degree of distortion,and can approximately restore the original image.Moreover,with the reduction of compression ratio and the increase of transmitted data volume,both the subjective quality and the objective indicators RMSE and PSNR are greatly improved.
作者
陈鑫
游敏娟
崔艺君
张勇
王世刚
CHEN Xin;YOU Minjuan;CUI Yijun;ZHANG Yong;WANG Shigang(School of Radiology,Shandong First Medical University(Shandong Academy of Medical Sciences),Tai'an 271016,China;Shandong Medicine Technician College,Tai'an 271016,China)
出处
《中国医学物理学杂志》
CSCD
2020年第12期1525-1528,共4页
Chinese Journal of Medical Physics
基金
国家级大学生创新创业训练项目(201810439058)
山东省自然科学基金面上项目(ZR2017MH078)
泰安市科技发展引导计划项目(2017NS0138)。
关键词
图像
嵌入式编码
离散余弦变换
均方根误差
峰值信噪比
image
embedded coding
discrete cosine transform
root-mean-square error
peak signal noise ratio
