基于自适应多方向稀疏模型的量化噪声均衡化图像解码方法

    Image Reconstruction Based on Compressive Sensing With Equalization Quantization Noise Model and Multi-direction Model

    • 摘要: 针对量化带来的非一致噪声,基于压缩感知理论,建立了基于量化噪声均衡化模型的图像优化解码方案.针对图像信号纹理的多方向特征,构建了自适应多方向稀疏表示模型.实验结果表明:以变换系数为观测数据,通过基于自适应多方向稀疏模型的量化噪声均衡化解码方法可以较大幅度地提高图像重建质量.

       

      Abstract: An error estimate method based on equalization quantization noise model was proposed for nonuniform quantization noise in image codec. With designed equalization matrix,a norm constraint which could enhance the quality of CS recovery significantly was shown. Furthermore,for directional textures of images,a multi-direction operator which could enhance the quality of CS optimize recovery was shown. A CS-based JPEG decoding scheme based on quantization error estimate and multi-direction operator was also presented,and experimental evidence exhibits more gains over CS reconstruction without error estimation and multi-direction operator.

       

    /

    返回文章
    返回