一种针对全景视频编码器的复杂度优化算法

相比于普通平面视频，全景视频能给用户提供更加沉浸式的体验，但其编码复杂度也急剧上升，阻碍了其推广应用。为解决此问题，本文首先结合全景视频纬度特性和帧间相关性构建复杂度预测模型，以更好地分配帧内复杂度资源；其次，构建一个转换模型将编码复杂度转换为时间复杂度，以确保帧内编码时间的稳定性；最后，综合上述两个模型提出一种自适应复杂度优化算法。实验结果表明，本文所提算法能在复杂度与编码质量之间取得更优的平衡，以适应不同应用场景的需求。

A complexity optimization algorithm for panoramic video encoder

Panoramic video provides viewers with more immersive experience than planar video. However, its coding computational complexity also increases dramatically, which is unfavorable to its widespread applications. To solve this problem, firstly, a complexity prediction model based on latitude characteristic and inter-frame correlation was proposed to optimally allocate intra-frame complexity resources. Then, a transform model, which transforms encoding complexity into time complexity, was presented to ensure the stability of intra-frame encoding time. Finally, by combining the above two models, an adaptive complexity optimization algorithm was proposed. Experimental results show that the proposed algorithm can achieve a better balance between computation complexity and compression quality to meet the different needs in diverse applications.