欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
对现有的DCT域恰可察觉失真（Just-Noticeable-Distortion，JND）模型进行改进，首先对JND模型中的亮度自适应因子进行准确化，其次在考虑纹理掩盖效应时引入平均边缘权重系数对图像进行分块，使JND模型更加符合人眼的视觉特性。并将改进的JND模型应用于图像编码中。实验结果表明，改进的JND模型和人眼特性的吻合性更好，在保证图像主观质量一定的情况下，与现有JND模型相比，可以容忍更多的失真，而基于改进JND模型的图像编码方法能够大大地提高编码效率。
英文摘要：
In this paper, an improved Just Noticeable Distortion （JND） profile in DCT domain is proposed to model the human vision system accurately and efficiently, which not only improves the luminance adaptation factor of JND profile, but also considers the texture masking effect based on block classification by the scheme of the mean edge-related weight of block. Furthermore, the proposed JND profile is applied on image coding. Experimental results show that the proposed model has a better consistent with human eyes characteristic and can tolerate more distortion comparing with the existing scheme at a given perceptual quality. The image coding scheme based on the proposed JND profile can achieve the higher coding efficiency.
同期刊论文项目
同项目期刊论文
Stereoscopic Visual Attention Based Regional Bit Allocation Optimization for Multi-view Video Coding
期刊信息
位置：