Multilayer image fusion method and system based on auto-focus technology

Abstract

The present invention provides a multilayer image fusion method and system based on the auto-focus technology. The method comprises: according to a fusion time and a hierarchical division on a processing layer of a fused input image, transforming the image from a spatial domain and a transform domain respectively; introducing the wavelet transform operation method into a convolution process, obtaining high-frequency information and low-frequency information after the transformed image is subjected to the lifting wavelet decomposition, and optimizing multiple thresholds in the fusion process; using the average structural similarity as a weight-optimized evaluation function in the fusion process; and performing high-frequency fusion on the high-frequency sub-band by using the image contrastcombined with the PCNN, and performing low-frequency fusion on the low-frequency sub-band by using the regional energy combined with the regional gradient. According to the technical scheme provided by the present invention, adaptive registration of high-resolution biological imaging can be implemented through an automatic and intelligent means, and the technical scheme has a simple operation, high detection efficiency, high reliability and strong versatility.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a multi-layer image fusion method and system based on autofocus technology. Background technique [0002] With the continuous development of computer technology, sensor technology and medical imaging technology, more and more medical images are successfully used in clinical diagnosis and treatment. Due to differences in sensor imaging principles, the information presented in different modal medical images of the same part or organ of the human body has its own characteristics, which brings new challenges to multimodal medical image fusion. Up to now, domestic and foreign scholars have not developed a complete set of theories and methods for pixel-level multi-modal medical image fusion, and have not established a unified framework suitable for fusion between different modalities. However, with the emergence of more modal medical images and the expansion of its wi...

AI Technical Summary

Problems solved by technology

However, different application fields have different requirements for images, and it is difficult to determine a unified and effective quantitative evaluation index at this time.

Images