Search

CN-116148347-B - Super-resolution imaging method for ultrasonic detection of internal defects of materials

CN116148347BCN 116148347 BCN116148347 BCN 116148347BCN-116148347-B

Abstract

The invention discloses a super-resolution imaging method for ultrasonic detection of internal defects of materials, which comprises the steps of obtaining an ultrasonic image, carrying out region segmentation on the image to obtain a plurality of small regions of N multiplied by N pixels, wherein the selection of the pixel size N is based on the inclusion of single defects as an original image, and only interpolation calculation is carried out on the defect regions in order to improve the timeliness of an algorithm. Determining the number of inserted pixel points by taking the representation ratio of an original image to an actual defect area as a basis, expanding the edge of the original image by adopting an edge filling method, selecting an adjacent 4X 4 image area at each inserted pixel point as an interpolation correlation area, and obtaining a pixel value at the pixel point to be inserted by space coordinate mapping, correlation area determination and distance weight calculation to finally obtain a super-resolution ultrasonic image, thereby improving the resolution of the original image. Compared with a cubic spline function interpolation method, the method simplifies algorithm complexity and improves imaging speed.

Inventors

  • SONG SHOUPENG
  • ZHENG LIN

Assignees

  • 江苏大学

Dates

Publication Date
20260512
Application Date
20220908

Claims (7)

  1. 1. The super-resolution imaging method for ultrasonic detection of the internal defects of the material is characterized by comprising the following steps of: 1) Carrying out ultrasonic phased array detection on the detected material to obtain an ultrasonic image of the region to be detected; 2) Region segmentation is carried out on the ultrasonic image of the region to be detected to obtain The pixels contain small areas of defects as the original image; 3) Determining the characterization proportion of the original image to the actual defect area according to the length and width of the defect image of the original image and the length and width of the corresponding actual defect area; 4) Expanding the edge of the original image by adopting an edge filling method to obtain an edge expanded image; 5) Image interpolation is carried out on the edge expansion image by utilizing a self-defined nonlinear interpolation function, the pixel coordinates of the super-resolution image are mapped to the edge expansion image through space coordinate transformation, the pixel values around the center point are obtained, the row-column distances between the coordinate points and the center point are respectively brought into the nonlinear interpolation function to form a weight matrix, each pixel point of the super-resolution image is calculated, and finally the super-resolution ultrasonic image is obtained; the nonlinear interpolation function in the step 5) The function expression is that, ; Wherein, the As a function argument, mapping the center point coordinate of the pixel point to be inserted into the super-resolution image onto the original image and the row-column distance of surrounding pixel points; The power is the power, and the value is a positive integer; And Is a function of the form factor, composed of The value range is limited.
  2. 2. The super-resolution imaging method for ultrasonic detection of defects inside a material according to claim 1, wherein said region segmentation in step 2) comprises the steps of: 2.1 Converting the ultrasonic image of the region to be detected into a gray image; 2.2 Setting an intensity threshold to divide the gray image background and defect information to form a binary image according to the distribution condition of the background and defect pixel values in the gray image; 2.3 Performing morphological closing operation on the binarized image to bridge the discontinuities and elongated gaps at the image defects, thereby smoothing the defect contour; 2.4 Acquiring coordinates of a defect contour boundary point of the smooth defect contour image; 2.5 Calculating the barycenter coordinates of the image defects according to the coordinates of the image contour boundary points; 2.6 Selecting centroid coordinates as center according to centroid coordinates of image defects The imaging area of the size serves as the original image.
  3. 3. The super-resolution imaging method for ultrasonic detection of defects inside a material according to claim 1, wherein the boundary filling method in step 4) selects pixel values of an edge row and a column of an original image, and expands the boundary of the original image by using a copy-edge-most pixel method to form an edge expansion image.
  4. 4. The super-resolution imaging method for ultrasonic detection of defects inside a material according to claim 1, wherein the mapping of spatial coordinates in step 5) specifically comprises: Assume that the coordinates of the pixel point to be inserted in the super-resolution image are Which maps to the center point coordinates on the original image The calculation is performed by the following formula, ; Wherein the method comprises the steps of And The scale is characterized for the super-resolution image to the defect region, The ratio of the number of super-resolution image lines to the number of original image lines, The ratio of the number of super-resolution image columns to the number of original image columns.
  5. 5. The method for ultrasonic detection of defects inside a material according to claim 1, wherein the determining of the associated region in step 5) comprises: Computing a perimeter mapped to a center point on the edge-expanded image Image region pixel values whose coordinates can be defined by 4 rows of coordinates And 4 column coordinates Determining that the calculation formula is that, ; Wherein, the , Representing the coordinates of a center point on the original image The numerical value is rounded down and is rounded up, And The number of lines and the number of columns of the edge expansion of the original image are respectively represented; Determining pixel values of 16 coordinate points around center point in edge expansion image, and pixel value matrix thereof In order to achieve this, the first and second, ; Wherein, the The image pixel values are extended for the edges, and the subscripts represent the corresponding row and column numbers.
  6. 6. The super-resolution imaging method for ultrasonic detection of internal defects of a material according to claim 1, wherein the distance weight calculation in step 5) specifically comprises the steps of: s5.1) calculating the pixel center of the point to be interpolated on the edge expansion image to be away from the periphery The distance between the pixels in the image area can be respectively 4 row distances And 4 column distances Determining that the distance calculation formula is that, ; Wherein, the The numerical value of (2) is located at In between the two, The numerical value of (2) is located at Between them; s5.2) distance 4 rows And 4 column distances Is correspondingly substituted into a nonlinear interpolation function to obtain Weight value matrix of distance between image area and center point to be interpolated And The expression is that, 。
  7. 7. The super-resolution imaging method for ultrasonic detection of defects inside a material according to claim 1, wherein the calculating of each pixel point of the super-resolution image in the step 5) comprises the steps of: calculating a pixel value of each pixel point to be inserted in the super-resolution image : ; Pixel values of pixel points to be inserted The super-resolution ultrasonic image is formed by inserting the ultrasonic image into the original image.

Description

Super-resolution imaging method for ultrasonic detection of internal defects of materials Technical Field The invention relates to the field of ultrasonic imaging, in particular to a super-resolution imaging method for ultrasonic detection of internal defects of materials. Background Ultrasonic detection technology is an effective and widely used non-destructive method in the industry. The visualization of ultrasonic detection data has become an important means for evaluating materials or structures, and how to improve the resolution of images has important value on the basis of the physical constraints of the existing detection and imaging methods, so that the method has self-evident meaning for quantitative evaluation of internal defects of materials. Therefore, with the development of computer technology, the requirements of people on ultrasonic images are higher and higher, and the fundamental limitation on resolution in the existing ultrasonic imaging is an urgent need of the modern ultrasonic detection technology. The current methods for improving the quality of ultrasonic images are mainly divided into two categories, namely a pretreatment technology and a post-treatment technology. Preprocessing techniques improve and optimize signal generation or image acquisition, including spatial and frequency compounding, harmonic imaging, pulse inversion, and the like, in relation to the physical characteristics of the signals involved, such as coherence, bandwidth, nonlinear propagation, attenuation, absorption, and the like. Post-processing techniques are methods that use image processing and machine learning (e.g., interpolation, bayesian analysis, antialiasing, deep learning, etc.) to enhance the image after the low resolution image is obtained. The image interpolation technology has wide application in an image processing system, and the image interpolation method can be adopted to acquire a high-resolution image from a single low-resolution image. Image interpolation methods are numerous and continuously developed, and the traditional interpolation algorithm is represented by a nearest neighbor interpolation method, a bilinear interpolation method, a bicubic interpolation method and the like. The nearest neighbor interpolation method takes the pixel value corresponding to the position to be inserted as the known pixel value nearest to the point. The nearest neighbor interpolation method is simple and quick to calculate, but is easy to generate sawtooth edges and mosaic phenomena. The bilinear interpolation method determines the weight of each pixel according to the distance between the corresponding point and the surrounding 4 adjacent points, and can obtain the pixel value of the target image. Compared with nearest neighbor interpolation, the bilinear interpolation method has smoother interpolation effect and no mosaic phenomenon, but detail information is easy to lose in edge detail processing. Furthermore, for nearest neighbor interpolation and bilinear interpolation, the basis function is a direct interpolation basis function, which means that the interpolation coefficient can be reduced to a pixel value. The bicubic interpolation method adopts a polynomial function, obtains the value of the pixel point to be interpolated by using the weighted average of adjacent sampling points around the interpolation point of the interpolation function, and needs to use a polynomial interpolation cubic function in the horizontal direction and the vertical direction respectively, so that smoother image edges can be created after interpolation processing, the processing result is finer, but the calculated amount is increased, and the time required for processing is long. Disclosure of Invention In order to solve the problems in the prior art, the invention provides a super-resolution imaging method for ultrasonic detection of internal defects of materials, and designs a brand-new nonlinear interpolation function, so that algorithm complexity is simplified, and image imaging speed and quality are improved. In addition, in consideration of the interrelationship and influence among pixels, the super-resolution image pixel calculation is performed by a plurality of pixels in the comprehensive area, and the effect is more excellent. The method has the characteristics of simple interpolation basis function structure, easy realization, easy parameter adjustment, improved image detail part and spatial resolution, and the like. In order to achieve the above object, the present invention provides a super-resolution imaging method for ultrasonic detection of internal defects of a material, comprising the steps of: 1) Carrying out ultrasonic phased array detection on the detected material to obtain an ultrasonic image of the region to be detected; 2) Performing region segmentation on an ultrasonic image of a region to be detected to obtain a small region with N multiplied by N pixels containing defects, wherein the small region