Research On Mass Image Database And Implementation

Mass data is referred as the very large amount of data, especially multimedia data and data from Internet. With the development of computer technologies, databases are very popular and the mass data is a big problem especially for the large images such as a TM image of hundred of millions and so on. The traditional technologies related to database can not deal with the mass data. It's a tough practical issue how to storage, manage and even to retrieval and to display the mass image data.The thesis tried to process the unlimited data utilizing the restricted memory and thus to storage and display the mass image data quickly. The pyramid model based on the wavelet analysis was used to represent the large image hierarchically for efficient storage and fast display.The original images were analyzed bases on the wavelet transformation and the coefficients were used to form a series of sub-images to reflect the different details of the original ones. The display requests with different zoon factors were answered with corresponding represents of the images. As we know that wavelet is an excellent time frequency transform and can compress the image data successfully. EZW algorithm was studied carefully and implemented in the thesis.Based on the popular image class libraries, Cximage, the traditional image operations were implemented such as display, zooming and so on. In order to deal with the immense image data, the file reading was blocked. The EZW algorithm, pyramid model storage and image management software based on oracle database system were implemented.