Adaptive medical image fusion method based on non-sampling contourlet transform

Abstract

The invention relates to an adaptive medical image fusion method based on non-sampling contourlet transformation, which belongs to the field of image processing. First, the source image is processed by arithmetic mean filtering, and then the image is decomposed by the orthogonal 9‑7 wavelet filter and the pkva filter in non-sampling to obtain the low-frequency subband coefficients and the subband coefficients of each band-pass direction; and then the low-frequency subband coefficients The edge information maximum criterion is used to select the fusion low-frequency sub-band coefficients, and the adaptive PCNN model based on the visual neuron model is used to select and fuse the band-pass sub-band coefficients for each band-pass sub-band coefficient; finally, the final fusion image is obtained through the inverse transformation of NSCT . The algorithm of the invention is very effective and correct, and the fused image edge and space texture information are clear, the color distortion is small, there is no false contour phenomenon, and the feature information of the original image is well preserved.

Description

technical field [0001] The invention belongs to the field of image processing, in particular to a medical image fusion method based on a non-sampling contourlet transform (NSCT) adaptive pulse-coupled neural network (PCNN). Background technique [0002] Image fusion refers to the synthesis of information about images or image sequences of a certain scene obtained by two or more sensors at the same time or at different times, so as to generate a new, more comprehensive and accurate description of the scene. Image. [0003] With the rapid development of medical imaging technology, the image quality has been greatly improved. However, due to the different imaging principles of medical imaging technology, using one modality imaging technology alone often cannot provide enough information for doctors. It is usually necessary to fuse medical images of different modalities to obtain comprehensive and complementary information in order to understand the comprehensive information of...

AI Technical Summary

Problems solved by technology

2006: 420-424"Research shows that the Contourlet transform needs to downsample the image, which makes the Contourlet transform produce false contours due to the lack of translation invariance

Telecommunications Technology, 2003, 3: 21-24" shows that artificial neural networks have been widely used in image fusion, especially the pulse-coupled neural network formed by Eckhorn et al. The field of image processing is being widely studied, but the connection strength of traditional PCNN is usually constant, which greatly limits the automatic processing ability and universality of PCNN

Images