一种新的图像分场描述编码错误掩盖算法

Novel error concealment algorithm for field-description coded image

针对信道中数据丢失导致图像恢复质量严重受损问题,提出将方向插值引入到基于图像分场描述编码的错误掩盖算法。将每帧图像分为两场独立编码传输,丢失一场数据时,通过改进Sobel算子计算正确接收的另一场图像内部边缘,沿边缘方向插值恢复出丢失的场。实验结果表明,应用这种基于像素行交织的错误掩盖算法后,图像主观恢复质量和峰值信噪比均较现有方法有显著提高。算法计算复杂度低,与H.264编码标准兼容,适于实际工程应用。

This paper presented a directional interpolation method for error concealment of field-description coded image over error-prone channel. It coded each frame into two field descriptions and transmitted them independently. It supposed that one of the descriptions was lost, then used the other correctly received description for the concealment. By modifying the Sobel operator and calculating the edge of correctly received field image, it could select suitable pixels to interpolate the lost field. Experiment results show that the proposed algorithm outperforms current algorithms in subjective image quality and the peak signal-to-noise ratio. The proposed algorithm is easy to implement, compatible with H. 264 coding standard and suitable for practical application.