Research On Key Technologies And Applications Of Image Processing

With the rapid development of the information age,the mobile Internet has been extensively used,thus a large amount of data and information are generated because of recording social and production activities.The rational use of these data and information can effectively improve human labor productivity.Among them,visual information occupies most of the information used by people.Therefore,it is important for people with the storage,processing and use of these image data.People have developed different methods to process these image data owing to different needs or purposes.First of all,rich information of image data results in large storage space and long transmission time,which make necessary the compression technology of image data.Furthermore,in the process of acquiring image data,due to the influence of physical equipment or people's factors,the quality of image data can not satisfy people's needs,so it is necessary to use image repair technology to process the image data.The final image are presented to people for visual observation in a certain way,which requires the drawing processes of image data.Therefore,our work focuses on the following topics: image compression technology,image repair technology,and parallel image rendering technology:First,we proposed an image compression technique based on adaptive block octree color quantization(AB-OCQ).By extending the traditional octree algorithm,the color quantization of adaptive block is performed based on the image.Compared with the mainstream image compression technology,the AB-OCQ method has a better performance in compression ratio under the premise of ensuring image quality,and also has the feature of directly accessing pixel data in a compressed state.This feature allows this technology to draw images more efficiently and consume less memory.Second,we developed an improved example-based image repair algorithm,which can repair large area of damaged image data or remove unnecessary parts of the image data.This technique improves the fixed template window used by the original algorithm,increases the set of template windows,and repair images by dynamically selecting the best matching template window.Experiments have been carried out,showing that the proposed algorithm achieves better performance in repairing images in comparison with other example-based image repair algorithms.Finally,we presented a parallel image drawing technique based on block technology.In this approach,we divides the drawing area and switch the traditional drawing procedure to the drawing commands of the sub-region,which are followed by performing task-level parallel processing on the drawing commands of the drawing sub-region.As compared to the traditional serial image rendering technology,Our experimental results show that under the same software and hardware conditions,this technology has a better performance in the parallel processing capabilities of multi-core computers,speedup,and the rendering efficiency.In summary,we are presenting: 1)an image compression technology based on adaptive block octree color quantization can perform high-quality compression of image data and provide technical support for image data storage and fast access;2)an improved example-based image repair algorithm can provide a better performance in image repairing and offer technical supports for the next processing of the image data;3)the proposed parallel image rendering technology based on blocks,which is high-efficiency image rendering technology,can break through the bottleneck of image rendering efficiency of traditional serial processing and take advantage of the parallel processing capabilities of multi-core computers.