In wireless collaborative communication networks, the study of audio and video coding and decoding for communication focuses on improving the stability of the network and the efficiency of data transmission. The goal of the node selection algorithm used in this research is to minimise the risk of network disruption, which is achieved through a global search algorithm. In addition, the key to effective segmentation of moving objects within the High Efficiency Video Coding (H.264) compression domain lies in extracting and processing motion vectors from the data stream, and utilising the mean-shift clustering algorithm to improve the accuracy of segmentation. Experimental results show that the developed algorithm exhibits high efficiency and accuracy on standard test sequences, especially in the detection and segmentation of moving objects. In addition, the algorithm achieves a good balance between compression efficiency and video quality by optimising the coding process and adjusting the Group of Pictures (GOP) structure and quantisation parameters. Comparing the rate-distortion curves, it can be seen that the prioritised coding strategy significantly improves the PSNR values of the moving object region, while the peak signal-to-noise ratio (PSNR) values of the background and the overall image are relatively stable, and this differentiated coding strategy contributes to the optimisation of the overall video quality, highlighting the importance of the coding process for dynamic content. These findings not only emphasise the advances in codec technology, but also provide important technical support for audio and video transmission in wireless networks.