| Image compression plays a vital role in image transmission and storage．Thewavelet transform can effectively overcome the limitations of the traditional Fouriertransform． And it has received much attention in the field of imagecompression．Fractal image compression for an image whose classes have obvioussimilarities is easy to achieve a high compression ratio, and also a very importantimage compression method．The sub-images in different resolutions of the sameorientation have many similarities in the wavelet transform．And these similaritiesprovide an opportunity for the combination of wavelet and fractal．Presently therehas been a few works about the hybrid wavelet-fractal coder. The peaksignal-to-noise ratio is improved, but the visual effect is not ideal. To solve thisproblem, the maximally stable extremal regions detection method is introduced intothe hybrid wavelet-fractal coder. The new method will be more consistent withhuman visual characteristics. Try to improve the peak signal-to-noise ratio,meanwhile, try to obtain a higher image quality. The main results are as follows.First, the formula of weighed peak signal-to-noise ratio is proposed on the basisof the traditional subjective and objective evolution criteria. The new evaluationcriterion will make an organic combination of subjective and objective evolutioncriteria. The peak signal-to-noise ratio is improved, at the same time, and subjectiveimage quality is also good.Last but not least, an improved compression method with a hybridwavelet-fractal coder is proposed. According to the result of detection, theprocessing methods of the high-frequency part and low frequency part are improved,respectively. Compared with the traditional method, the selection of detailed imageblocks is more consistent with human visual system. The low frequency part isweighted to a reasonable increase. The experimental results show that the objectiveevolution and subjective quality of the reconstruction images are both good. |
PDF Full Text Request