针对聋哑人移动手语视频通信,提出一种H.264计算资源和比特资源联合优化分配方法。该方法根据聋哑人视觉选择特性将手语视频划分为不同的区域,通过为不同区域分配不同的量化系数进行比特资源优化分配;并将H.264手语视频编码复杂度分为三个级别,根据能量受限设备的电池能量自适应地选择编码级别进行计算资源优化分配。研究结果表明:该算法在保证手语视频感兴趣区编码质量的同时,可以有效降低编码器的计算复杂度,延长电池工作时长。
In order to realise the mobile sign language video communication for deaf people, we propose a joint optimised allocation scheme for H. 264computational resource and bit resource in this paper. According to visual selectivity characteristics of deaf people, the scheme divides the sign language video into different regions, and optimises the allocation of bit resource by allocating different quantisation parameters for different regions. Also, it ranks coding computational complexity of H. 264 sign language video in three levels, and optimises the computational resources allocation by adaptively selecting the corresponding coding complexity level on the basis of battery energy of resource-limited devices. Study results show that the proposed scheme can effectively reduce the computational complexity of encoder and prolong the running hours of battery while maintaining the coding quality of region-of-interest in sign langnage video.