This paper presents a Personalized large Social Image Transmission method in mobile wireless network(MWN) environment, called the PSIT. The whole transmission process of the PSIT works as follows: first, when a social image I S is prepared to transmit from a sender node to user U R , a preprocessing step is then conducted to obtain the optimal image fragment(IF) replica based on the users' preference model and the network bandwidth at the sender node. After that, the candidate IFs are transferred to the receiver node from the slave one according to the transmission priorities. Finally, the IFs can be recovered and displayed at the receiver node level. The proposed method includes five enabling techniques: 1) neighborhood-based tag enrichment processing, 2) user attention degree(UAD) derivation of the regions of interest(ROI), 3) an adaptive multi-resolution-based IF replica selection method, 4) a UAD-based IF replica placement method, and 5) a priority-based robust IF transmission scheme. The experimental results show that the performance of our approach is both efficient and effective, minimizing the response time by decreasing the network transmission cost while increasing the parallelism of I/O and CPU.