基于三维自组织映射的视频帧间预测编码算法  

Inter-frame predictive coding algorithm for video based on 3D self-organizing map

作　　者：覃远年[1] 陈超群 QIN Yuannian;CHEN Chaoqun(School of Information and Communication,Guilin University of Electronic Technology,Guilin 541004,China)

机构地区：[1]桂林电子科技大学信息与通信学院,广西桂林541004

出　　处：《中国高新科技》2025年第1期126-127,133,共3页

基　　金：广西自然科学基金(桂科2020GXNSFAA159146);广西创新驱动发展专项(桂科AA21077008);“认知无线电与信息处理”教育部重点实验室基金项目(CRKL190108)。

摘　　要：文章针对视频帧间编码中运动估计和运动补偿(ME+MC)算法的不足,提出一种基于三维自组织映射的帧间预测编码方法。该方法将FS-3DSOM算法用于运动模式库的训练,通过运动模式库和运动估计算法分别获取最佳匹配块,选择率失真代价最低的匹配块作为预测块,从而提高帧间预测准确度。同时,为了减小运动模式库的匹配误差,采用了基于聚类的运动模式库构造方法。实验结果表明,在相同码率的情况下,基于三维自组织映射的帧间预测编码方法比传统方法具有更好的编码性能。Aiming at the shortcomings of motion estimation and motion compensation(ME+MC)algorithms in video inter-frame coding,an inter-frame predictive coding method based on 3D self-organizing map is proposed.The method uses the FS-3DSOM algorithm for the training of the motion pattern library,obtains the best matching block by the motion pattern library and the motion estimation algorithm respectively,and selects the matching block with the lowest rate-distortion cost as the prediction block,to improve the video coding performance.Meanwhile,in order to reduce the matching error of the motion pattern library,a clustering-based motion pattern library construction method is adopted.The experimental results show that the inter-frame predictive coding method based on 3D self-organizing map has better coding performance than the traditional method at the same bit rate.

关 键 词：自组织映射 视频编码 帧间预测 运动估计 运动补偿 

分 类 号：TN919[电子电信—通信与信息系统]