Microscope image processing method

Abstract

The invention discloses a microscope image processing method. The method comprises the following steps: extracting an image block with effective information from a microscope image; training a generative adversarial network to reconstruct an image block with effective information, including training a generative adversarial network model, storing the trained model and reconstructing the image block with effective information into a reconstructed image block with effective information; and splicing the reconstructed image blocks with the effective information to form a processed microscope image. According to the microscope image processing method disclosed by the invention, an integral solution for solving the super-resolution problem of the medical microscope image is provided, the network training and whole microscope image reconstruction efficiency can be improved, reconstruction errors can be minimized, work efficiency of operators (such as doctors) can be effectively improved, andrelatively strong robustness and robustness are realized.

Description

technical field [0001] The invention relates to the field of image processing, and more particularly relates to a method for processing microscope images. Background technique [0002] As we all know, images are one of the most important information carriers in our daily life. However, in the process of generating images, due to the influence of hardware equipment and the external environment, the obtained images often cannot meet people's requirements, resulting in certain image cognition obstacles. Therefore, how to improve the resolution of images has become a hot issue in the field of image processing, and therefore, the technical research of image super-resolution has attracted extensive attention from all walks of life. [0003] At present, the method of deep learning for super-resolution research of ordinary images has achieved breakthrough results. However, the spatial resolution of microscope images is limited by the Abbe diffraction limit, as well as the influence...

AI Technical Summary

Problems solved by technology

[0003] At present, the method of deep learning for super-resolution research of ordinary images has achieved breakthrough results. However, the spatial resolution of microscope images is limited by the Abbe diffraction limit, as well as the influence of illumination and noise. difficulty

It is understood that the current ordinary microscope equipment, the maximum magnification of images that can be collected by skilled personnel is 4.5, which cannot provide users, such as doctors, with higher resolution images; and due to the limitation of the microscope's viewing area, When the image is enlarged, the complete image information cannot be collected. If a complete enlarged image is required, we need to stitch and fuse the collected images. In this case, it is also difficult to ensure the complete collection of image information.

Images