Mass Image Data Management System

With the development of computer technology, multimedia technology andnetwork technology, especially the development in the field of network technology,data collection technology and image processing technology, which closely related tocomputer application technology, brought about the improvement of the degree ofindustry information. In the large enterprise and goverment application systems,multimedia information processing has been an indispensable part of the system.While the conversion from paper documents in the office information processing toelectronic document has also become the trend of improving information management.This triggers the data scale and range of application of the storage, processing anddata services expanding gradually. In order to support a large-scale image data storageprocessing applications, development of large-scale image data processing to adapt tothe professional management system for solving massive image data storage, retrieval,processing, and service efficiency is a very important way. Based on the aboveapplication requirements, our scientific research group developed the mass datastorage platform, the main work in the following areas:(1) A massive image data storage structure model is designed and implemented. Inthe underlying it uses table store metadata, stores attribute data with the block,and uses the data packet in the operating system layer organize hierarchy, which isindependent from the operating system and hardware system for storage,organization and management of massive image data.(2) The use of the data cache data optimize data access structure, reduce data readingand writing disk access times in the case of large-scale, especially for metadataaccess, and improve the overall system performance.(3) Data retrieval modifies the combination of dynamic hashing and B+tree, whilerandom access uses the hash table, and range query uses the B+tree.(4) Design and implement an efficient data access services. For multi-user access tomulti-task state, using multi-thread management achieve full use of hardwareresources.(5) It has abundant image processing functionalities and was developed andcompleted the image processing toolbox in the way of plug-in. For document image processing features, which makes it practical as image processing tools andconvenient for the people who use the system for image processing.