Particle morphology recognition method based on interference imaging and Fourier transform

By using the method based on interference imaging and two-dimensional Fourier transform, the interference defocus image is processed to identify the particle morphology in the multi-dispersed particle field, which solves the problem of poor recognition effect in the prior art and realizes the multi-to-multiple Accurate morphological recognition and microphysical properties measurement of dispersed particles.

CN118913134APending Publication Date: 2024-11-08TIANJIN POLYTECHNIC UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410963859.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The prior art has failed to effectively use the two-dimensional Fourier transform and the spectrum of the interference defocused image to identify particles in the polydispersed particle field, especially the recognition effect of non-spherical particles, and the existing methods are insufficient in real-time. .

Method used

Using particle morphology recognition methods based on interference imaging and two-dimensional Fourier transform, the interference defocus image is enhanced by contrast, median filtering, low-pass filtering and binarization, combined with the traversal and marking of white pixels, the white is counted. The overall number of pixels is used to distinguish spherical and non-spherical particles.

Benefits of technology

The morphological recognition of particles in the polydispersed particle field is realized, providing accurate microphysical characteristic measurement support for the polydispersed particle field, which can effectively distinguish spherical and non-spherical particles, and improve the accuracy and real-time recognition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure HDA0004951150170000011
    Figure HDA0004951150170000011
  • Figure HDA0004951150170000012
    Figure HDA0004951150170000012
  • Figure HDA0004951150170000021
    Figure HDA0004951150170000021
Patent Text Reader

Abstract

The invention discloses a particle morphology recognition method based on interference imaging and Fourier transform. The method comprises the following steps: step 1, establishing an interference particle imaging system, and obtaining a single-particle interference defocus image of a multi-dispersion system particle field; step 2, contrast enhancement and median filtering time domain denoising are carried out on the obtained interference defocused image; step 3, performing two-dimensional Fourier transform on the processed image, and performing low-pass filtering after a frequency spectrum is obtained; step 4, carrying out binaryzation on the frequency spectrum after low-pass filtering; and step 5, traversing white pixel points in the binary image, if white pixel points still exist in eight neighborhoods around the white pixel points, counting the overall number of the white pixel points, if the number of the white pixel points is 1, determining that the particles are non-spherical particles, and if the number of the white pixel points is greater than 1, determining that the particles are spherical particles. According to the invention, identification of particle morphology in a polydisperse particle field can be realized, and technical support is provided for measurement of a complex particle field.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of particle morphology measurement and image processing in an optical system, and in particular to a particle morphology recognition method based on interference imaging and Fourier transform. Background Art

[0002] Interference particle imaging technology has the advantages of high precision, wide measurement range and non-contact, so it is widely used in the field of particle measurement. During the measurement process, the interference focused image of spherical particles is a two-point image, and the interference defocused image is a stripe image with light and dark alternating; the interference defocused image of non-spherical particles is a speckle image. The interference defocused image of spherical particles has three bright spots in the center of the frequency domain, and there are three point-like white pixel areas after binarization. The interference defocused image of non-spherical particles has multiple bright spots in the center of the frequency domain, and there is only one block-like white pixel area after binarization. The morphological information of the particles can be obtained by analyzing the number of white pixel areas after the interference defocused image spectrum is binarized. Therefore, converting the interference defocused image to the frequency domain for analysis is of great significance for particle morphology recognition.

[0003] At present, there is no method in the prior art to identify the morphology of particles in a polydisperse particle field using a two-dimensional Fourier transform and based on the spectrum of the interference defocused image. The morphology recognition method of a polydisperse particle field currently has a particle morphology recognition method based on polarization characteristics, but this method is mainly aimed at spherical particles and ellipsoidal particles, while the range of non-spherical particles is very large, and ellipsoidal particles cannot cover all non-spherical particles well. In addition to this method, there is also a particle morphology recognition method based on improved YOLOv7, but this method is a morphology recognition under multi-particle imaging, and the real-time performance is poor. Summary of the invention

[0004] On the basis of the above-mentioned prior art, the present invention proposes a particle morphology recognition method based on interference imaging and Fourier transform. By performing a two-dimensional Fourier transform on the interference defocused image and binarizing the spectrum, the white pixels are traversed and the number of white areas is obtained through discrimination. The method is suitable for the morphology recognition of particles in a polydisperse particle field.

[0005] The technical solution adopted by the present invention is: a particle morphology recognition method based on interference imaging and Fourier transform, which is carried out according to the following steps:

[0006] Step 1: Build an interferometric particle imaging system to obtain a single-particle interferometric defocused image of a polydisperse particle field;

[0007] Step 2: Perform contrast enhancement and median filtering on the obtained interference defocused image in the time domain to remove noise;

[0008] Step 3: Perform a two-dimensional Fourier transform on the processed image, obtain the spectrum and then perform a low-pass filter;

[0009] Step 4: Binarize the spectrum after low-pass filtering to obtain a binary image;

[0010] Step 5: Traverse the white pixels in the binary image. If there are white pixels in the eight neighborhoods around the white pixels, mark these white pixels as a whole. Count the number of marked white pixels in the traversal result. If the number is one, it is a non-spherical particle. If the number is greater than one, it is a spherical particle.

[0011] The present invention has the following beneficial effects:

[0012] The present invention can realize the particle morphology recognition of the polydisperse particle field and provide technical support for the measurement of other microphysical properties of particles in the polydisperse particle field. The polydisperse particle field is composed of different types of particles, which can be divided into spherical and non-spherical particles. Non-spherical particles also have various shapes, such as linear, branch-like, plate-like, etc. The microphysical properties of the particles include phase state, size, refractive index, etc. Only by determining whether the particle is spherical or non-spherical through morphology recognition can the phase state of the particle be determined as a liquid phase particle or a solid phase particle, and other microphysical properties such as size, refractive index, shape, etc. can be accurately measured and obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a flow chart of the polydisperse particle morphology identification method of the present invention;

[0015] Figure 2 This is the structure diagram of the experimental device of the interferometric particle imaging system;

[0016] Figure 3 Schematic diagram of the interference defocused image and processing results. DETAILED DESCRIPTION

[0017] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0018] Implementation steps: Interferometric particle imaging system

[0019] The interference particle imaging system used in the present invention is as follows Figure 2As shown, the laser 1 emits a laser beam with a wavelength of 532nm and a maximum output power of 3W. The laser beam is expanded and filtered by a spatial filter composed of a microscope objective 2 and a pinhole 3. After being collimated by a collimating lens 4, the beam is divided into two beams by a 1:9 beam splitter. One beam is calibrated by a reflector to calibrate the imaging lens and CCD (covered during the experiment), and the other beam is compressed into a 1mm sheet laser beam by a plano-convex cylindrical lens 5 and a plano-concave cylindrical lens 6. The imaging lens 8 for collecting the scattered light of the particles is a Nikon 50mm f / 1.4D fixed-focus lens, the receiving device 9 is a CCD with a pixel size of 3.45μm and an effective pixel number of 2448×2048, the particles to be measured are polystyrene spherical particles with particle sizes of 45μm, 60μm, and 90μm, and sand particles with a size of 170-280μm. In the interference imaging of mixed particles, a syringe is used to absorb polystyrene spherical particles diluted with alcohol, and then tweezers are used to clamp the sand particles and place them on a glass slide 7 for imaging. The system parameters are: object distance is 100mm, defocused image distance is 110mm, and the system magnification is 90°.

[0020] according to Figure 1 In the steps shown, firstly, the interference defocused images of spherical and non-spherical particles of different particle sizes are obtained through the IPI system. Since the background of the defocused image is dark and the brightness of the stripes or speckles is low, the contrast of the image is enhanced. There is impurity interference in the experimental system, and the defocused image will have noise. The defocused image is denoised in the time domain, and the filter used is a median filter. After denoising, the image is subjected to a two-dimensional Fourier transform. Since the difference point between the spectrum of spherical particles and non-spherical particles is located in the low-frequency part of the center of the spectrum, there is little difference between spherical and non-spherical particles in the surrounding high-frequency parts, and the high-frequency part has great interference with the subsequent binarization operation, the spectrum is low-pass filtered, and the spectrum is binarized after low-pass filtering. The threshold selected is 0.8 times the maximum value of the spectrum pixel. Figure 3 The figure shows the schematic diagram of the interference defocus image and the processing results. Among them: (a)-(d) are the interference defocus images of spherical and non-spherical particles of different sizes; (e)-(h) are the results of contrast enhancement of the interference defocus image; (i)-(l) are the results of time domain denoising of the interference defocus image after median filtering; (m)-(p) are the spectra of the interference defocus image after two-dimensional Fourier transform and low-pass filtering; (q)-(t) are the binarization results of the filtered spectrum. By traversing the white pixels in the binary image, if there are white pixels in the eight neighborhoods around the white pixels, they are marked as a whole, and finally the number of marked white pixels is counted. If the number is equal to one, it is a non-spherical particle, and if the number is greater than one, it is a spherical particle.

Claims

1. A particle morphology recognition method based on interferometric imaging and Fourier transform, It is characterized in that The method comprises the following steps: Step 1: Build an interferometric particle imaging system to obtain a single-particle interferometric defocused image of a polydisperse particle field; Step 2: Perform contrast enhancement and median filtering on the obtained interference defocused image in the time domain to remove noise; Step 3: Perform a two-dimensional Fourier transform on the processed image, obtain the spectrum and then perform a low-pass filter; Step 4: Binarize the spectrum after low-pass filtering to obtain a binary image; Step 5: Traverse the white pixels in the binary image. If there are white pixels in the eight neighborhoods around the white pixels, mark these white pixels as a whole; count the number of marked white pixels in the traversal result. If it is 1, it is a non-spherical particle, and if it is greater than 1, it is a spherical particle.