Fractal image compression coding method based on CNN and GEP

Abstract

The invention discloses a fractal image compression coding method based on CNN and GEP, which comprises the following steps: respectively segmenting an image to be compressed into a plurality of valuedomain sub-blocks which are not overlapped and a plurality of overlapped definitional domain sub-blocks, forming a value domain block pool by the plurality of value domain sub-blocks, and forming a definitional domain block pool by the plurality of definitional domain sub-blocks; utilizing the optimal convolutional neural network model to classify the value domain block pool and the definitionaldomain block pool respectively to obtain a corresponding value domain block classification pool and a corresponding definitional domain block classification pool; and calculating a transformation parameter corresponding to each value domain sub-block set in the value domain block classification pool and a definitional domain sub-block set of the same category in the definitional domain block classification pool by utilizing an optimal gene expression programming model, so as to realize fractal image compression coding. According to the method, the compression encoding speed can be greatly increased, the encoding time is shortened, a high compression ratio is obtained, and the actual use requirement can be met.

Description

technical field [0001] The present invention relates to an encoding method. More specifically, the present invention relates to a fractal image compression coding method based on CNN and GEP, which belongs to the technical field of fractal image compression coding. Background technique [0002] At present, image compression standards mainly adopt technologies such as discrete cosine transform (DCT) and wavelet transform (DWT), which are mature, but the compression ratio is not high. The fractal image compression technology is a completely different coding method from the traditional compression technology, and it is mainly realized by fractal self-similarity and iterative function system IFS. Barnsley applied the IFS (Iterated Function System) theory to image compression coding for the first time, and obtained a very high compression ratio, but the coding process required manual participation. Jacquin proposed a scheme based on the block iterative function system (PIFS), t...

AI Technical Summary

Problems solved by technology

However, the huge amount of calculation still restricts the speed and time of encoding

For this reason, people have proposed a parallel algorithm for fractal image compression on a variety of parallel platforms, which greatly improves the compression coding speed and reduces the coding time, but it still does not meet people's actual needs.

Images