RIA: Vector Quantization for Image Compression and Processing

点击翻译按钮获取中文摘要

Image compression is the process of reducing the amount of data required to store or transmit an image. It has traditionally been done with little regard for image processing operations that may precede or follow the compression step. It is proposed to investigate areas in which image compression and image processing may be combined into a single operation, and also areas in which image compression may lead to improved or simpler image processing operations. Vector quantization algorithms for the compression of halftoned binary images will be developed and images will be compared that are first compressed and then halftoned with images that are halftoned and then compressed. Vector quantizers that can simultaneously perform both compression and image processing tasks such as halftoning, inverse Fourier transforming, and contract enhancement will also be investigated. This will lead to a reduction in computational complexity if the complexity of the compression step is less than that of the image processing operation. Finally, examinations will be done to see if image compression will lead to simpler or improved image processing operations such as image segmentation, target detection, and pre- screening for suspect regions in medical images.