The Generation Of Components Of Images And The Research Of Interlayer Compression Method

In recent decades,the ideas of new requirements for images and video streaming to adapt to the the diversity of network terminal and the complexity of the equipment put forward along with the rapid development of network and multimedia technology,providing different resolution and quality of images became the inevitable trend of network development. In this case,scalable coding technology arises at the historic moment. In the process of dealing with big data image,spatial scalable coding has been widely used and how to deal with the relation of resolution and quantity of the image has become the focus.In order to deal with the compression and display of large data image effectively,there is an effective method to compress the data by applying the pyramid model. In the pyramid model, the resolution reduced step by step until the top layer, it forms the next layer by reducing ratio of the original image. And in the process of building the pyramids, how to carry out the sampling and the compression mode becomes the key. It used double linear and double three inferior to generate the pyramid model in the conventional method, and used JPEG method to compress the data, but the resolution of the generated image needs to improve and the compression technology is also need to be furtheroptimized.As the users’ requirements for the number and quality of software, the traditional software development methods have been unable to meet demand, so the software component becomes a trend of development. Component is a kind of encapsulation of the combination of classes to complete the functions of the specific services, and it defines a series of interfaces, which only communicates with the outside world through the interfaces to provide services. The implementation of component can improve the efficiency of the use of the software,and it’s also convenient and be assembled. But how to effectively create a software component is also become a problem to study.Based on two aspects of the problems above, the pyramid model designed in this project greatly improves the quality of the image based on the research of optimization of the traditional method to improve resolution, and increased the application of component in the process of compression to design a new compression method which greatly reduce the size of image. In this dissertation,the main research work is as follows:Put forward a new scalable coding process about the compression method among layers and generation of the big data image. In the generation of the image data, fully analysed the prediction and searched process of AVS coding standard to get the texture information of image which means the directions of change in the image, and then according to the different directions to choose different filters in the down-sampling,this method fully reserves the informationof the upper layer and gets higher resolution of each layer newly generated.Made it into a atomic component with the down-sampling process.When dealed with the process of big data image,then divided the image into the images with the same size, and these segment images invoke the atomic component and also divided a picture of a high resolution into many small images to deal with at the same time which highly improves the speed. In the term of the compression of image between the layers, only compressed the residual,not encode each layer compression like the traditional compression methods,the method greatly reduce the data redundancy and improve the compression ratio. In the coding process, get the residual by the result from the subtracting of the down-sampling in this layer and the up-sampling of the refactoring data. Because of the small residual data, the dissertation counts and analyses the data after DCT transformation and by modifying and optimizing the quantitative process, the image size is greatly reduced after entropy coding.Making it encapsulation as atom components of the IDCT, inverse quantization and entropy decoding process in decoding, these components fully applied to the entire compression process of the data between layers.