Rational interpolation zooming method based on CT image edge

Abstract

The invention discloses a rational interpolation scaling method based on the edge of a CT image, which comprises step 1, obtaining the boundary of a single pixel of the image through an edge detection algorithm; step 2, obtaining a fitting curve of the boundary pixels, and resampling to obtain a new enlarged boundary; Step 3, resampling non-boundary pixels according to different situations. The boundary curve and key pixels are obtained through the edge detection algorithm, the fitting curve of the boundary pixels is obtained when zooming in, and the new boundary is obtained by resampling to maintain the sharpness of the boundary visually; non-boundary pixels are divided into rational interpolation and ordinary according to different situations There are two types of bilinear interpolation. Due to the special processing of the image edge, the loss of edge information during the image scaling process is avoided, so that the details of the image can be better displayed, and the image information of the lesion can be displayed better. The magnified CT image can clearly show the relationship between the lesion and the surrounding normal tissues and organs, so that some lesions that are not easy to be found can be detected early.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a rational interpolation scaling method based on CT image edge features. Background technique [0002] Image scaling is a basic problem in image processing and has been widely used in practice. For example, image scaling is often used in medical, aerospace, and video applications. Its main purpose is to make the new image better reflect the original image. The shape of the scene, especially the detail information. In particular, in medical diagnosis, the observation of CT images, which is an important auxiliary diagnostic means for clinical disease diagnosis, often requires zoom-in operation so as to be suitable for details, especially the observation of lesions, so as to improve the accuracy of doctor's disease diagnosis. [0003] The essence of image magnification is to generate gray values ​​of unknown high-resolution pixels based on known gray values ​​of low-resolu...

AI Technical Summary

Problems solved by technology

In addition to hard defects such as medical hardware equipment and radiation dose limitation, the edge between lesion tissue and normal tissue in CT images is still relatively blurred. If

Images