Design And Implementation Of Matching Technical Solution For High Performance Image In Mobile Phone Communication

With the improvement in the bandwidth and quality of mobile network, multimedia message has become the major carrier of multimedia information transmission. However, network development has also left many technologies lagging behind. In China Mobile’s message service, the recognition and comparison of multimedia messages is one of the most important functions that suppliers are providing. The development of technologies for image recognition and comparison that can satisfy telecommunication-level application is therefore imminent.Based on the core processing module of China Mobile’s "Multimedia Message Tracking Platform", this Study aims to realize image comparison and analysis. The current study results are still time-consuming in feature extraction, comparison and analysis due to the complex nature of image itself, and can hardly satisfy the need for telecommunication-level application (China Mobile requires that the maximum recognition time for a multimedia message of 100K should be no more than two seconds).So it never used in the field of telecommunications applications. Therefore, the establishment of a platform that can conform to telecommunication-level standard is of great significance to the expansion of multimedia message service.Following studies on image recognition algorithm ideas, the main goal of this thesis is to design a new, more efficient MMS picture detection algorithm, demonstrate the feasibility of the algorithm and to verify the algorithm accuracy with experiment. The system uses LINUX C as development language, has good stability and scalability. Aiming at the existing defect characteristics of MMS image recognition and traditional Hash images, this thesis proposes a suitable for MMS picture recognition framework, provide effective solutions for multimedia message service. The scheme not only keeps the traditional Hash method for perceptual robustness and low collision probability, but also achieves the telecom level requirements in performance on storage optimization and alignment of image database. New, the algorithm has been applied to the actual project, and implemented in the cooperation project between Neusoft and China Mobile. It achieved good results and has been highly praised by users.