Search

CN-121067764-B - System and method for detecting surface characteristics of nano silver brush hair

CN121067764BCN 121067764 BCN121067764 BCN 121067764BCN-121067764-B

Abstract

The invention discloses a system and a method for detecting the surface characteristics of nano silver bristles, which belong to the technical field of visual detection, and solve the problems of dead zone and single index of curved surface detection by acquiring and splicing full-surface images of the bristles through a linear array camera and a rotary carrier, extracting gray image analysis uniformity. The method comprises the steps of detecting the penetration depth of bristles with qualified uniformity, capturing friction shedding particles by using a fluorescent marker, constructing a distribution image sequence, triggering damage imaging by combining a roundness threshold value and similarity analysis, quantifying a damage area to track the abrasion evolution of a coating, immersing the bristles in a simulation solution containing a fluorescent probe in an abrasion stabilization stage, dynamically sampling and analyzing fluorescent data slope, intelligently distinguishing burst release of agglomerated particles from stable release of monodisperse particles, and accurately evaluating a release process, thereby solving the problems of incomplete uniformity evaluation, insufficient abrasion mechanism analysis and inaccurate release performance detection in the conventional detection.

Inventors

  • LIU WENZHI
  • LI MEIJUAN
  • ZHAN RUI
  • WANG YOUFA

Assignees

  • 湖北瑞特威科技股份有限公司

Dates

Publication Date
20260512
Application Date
20250818

Claims (8)

  1. 1. Nano silver brush hair surface property detecting system, its characterized in that includes: Collecting linear images covering bristle samples, extracting and screening characteristic points of adjacent linear images, positioning nano silver particles through template matching and Gaussian fitting after calculating a homography matrix, obtaining offset of the adjacent linear images, splicing to generate a full-surface image of the bristle samples, extracting a bristle main body area of the full-surface image, converting the bristle main body area into a gray image, and judging whether macroscopic uniformity of the gray image is qualified or not through initial quantitative detection of distribution uniformity of nano silver on the surface of the bristle samples; If the macroscopic uniformity is judged to be unqualified, dividing the severity of the defect, otherwise, judging the macroscopic uniformity to be qualified, and entering a depth detection link, wherein the depth detection link comprises friction loss detection and release detection; The friction loss detection is used for simulating a friction process of the nano silver brush hair, carrying out friction detection on a brush hair sample with qualified macroscopic uniformity, carrying out fluorescence imaging on a nano silver particle collecting device after single friction, constructing a nano silver particle distribution image sequence, triggering acquisition of a brush hair sample surface damage image and constructing a damage image sequence by constructing a similarity analysis model of adjacent particle distribution images; The release detection is used for immersing the rubbed bristle sample into a simulation solution containing a fluorescent probe when the friction loss detection is terminated, collecting fluorescence data of the sample solution, and detecting the release amount of nano silver particles of the rubbed bristle sample in the simulation solution; The specific steps of friction loss detection include: Carrying out fluorescent marking on the bristle sample with qualified macroscopic uniformity, collecting linear images of the bristle sample after fluorescent marking, and splicing to generate a bristle sample reference image; setting a friction working condition, starting a friction test bed to rub a bristle sample, and simultaneously placing a nano silver particle collecting device below a friction area to capture the fallen nano silver particles; After each brushing sample is rubbed for a single time, fluorescent imaging is carried out on the nano silver particle collecting device, nano silver particle distribution images are collected, and a particle distribution image sequence corresponding to the rubbing times is constructed; Performing binarization treatment on the particle distribution image to obtain a nano silver particle mask, and filling cavities in the nano silver particles through morphological operation; Configuring a roundness threshold, analyzing and marking all nano silver particles through a communication area, obtaining the mass center and the area of the nano silver particles, calculating the roundness of each nano silver particle, marking the nano silver particles larger than the roundness threshold as monodisperse particles, otherwise marking the nano silver particles as agglomerated particles; constructing a similarity analysis model of adjacent particle distribution images, extracting the difference of the number of the monodisperse and agglomerated particles and the mask deviation degree of the nano silver particles as characteristic parameters, and obtaining the comprehensive similarity of the adjacent particle distribution images through weighted average; Configuring a similarity threshold, when the comprehensive similarity of adjacent particle distribution images is smaller than the similarity threshold, collecting linear images of bristle samples, splicing to generate surface damage images of the bristle samples, and constructing a surface damage image sequence corresponding to the friction times; detecting damage areas of the surface damage images, configuring damage threshold values based on a reference image difference method, extracting areas with gray value changes larger than the damage threshold values as suspected damage areas, connecting the suspected damage areas through morphological operation, and quantifying the area occupation ratio and the space distribution of the suspected damage areas; Counting the periodic friction times of the last acquired surface damage image, and judging that the coating of the bristle sample is worn into a stable stage and stopping friction loss detection when the periodic friction times are larger than a preset stable detection threshold value; And configuring a limit frequency threshold and a limit area threshold, and stopping friction loss detection when the accumulated friction frequency is larger than the limit frequency threshold or the area occupation ratio of the suspected damaged area is larger than the limit area threshold.
  2. 2. The nano-silver bristle surface property detection system according to claim 1, wherein the specific step of obtaining the adjacent linear image offset comprises: Performing self-adaptive histogram equalization on adjacent linear images, extracting characteristic points of nano silver particles in the linear images, and limiting the searching direction of the characteristic points to be the circumferential direction perpendicular to the axis of the bristle sample; Configuring a space constraint threshold and a gray constraint threshold, calculating the angle difference between adjacent linear image feature point pairs in the circumferential direction and the gray similarity of the region where the angle difference is located, screening out feature points with the angle difference smaller than the space constraint threshold and the gray similarity larger than the gray constraint threshold, and marking the feature points as matching point pairs; Calculating a homography matrix of the matching point pairs, and carrying out sub-pixel level positioning on the nano silver particles in the overlapping area by adopting a template matching and Gaussian surface fitting algorithm based on the homography matrix; and positioning the corresponding area of the overlapping area in the adjacent linear images based on the homography matrix, calculating the corresponding point of the positioned nano silver particle point in the corresponding area through the homography matrix, obtaining the offset of the sub-pixel coordinates of the corresponding point, and obtaining the offset of the overlapping area of the adjacent linear images through average value calculation.
  3. 3. The system for detecting the surface characteristics of the nano-silver brush hair according to claim 2, wherein the specific step of performing sub-pixel level positioning on the nano-silver particles in the overlapping area by using a template matching and Gaussian surface fitting algorithm comprises the following steps: Establishing a unified coordinate system for adjacent linear images according to the homography matrix, carrying out homography transformation on the matching point pairs, calculating the overlapping area of the adjacent linear images and dividing the overlapping area into subareas; for each sub-region, matching the normalized cross-correlation coefficient with a nano silver template library, screening candidate sub-regions, taking the centers of the candidate sub-regions as candidate points, and constructing a candidate point set; Setting a neighborhood window, carrying out Gaussian surface fitting on each candidate point in the neighborhood window, configuring a fitting threshold, and marking as invalid candidate points and removing the invalid candidate points from the candidate point set if the decision coefficient of Gaussian surface fitting in the neighborhood window of the candidate points is smaller than the fitting threshold; otherwise, acquiring the extreme point of Gaussian surface fitting as the sub-pixel level coordinate of the nano silver particle at the candidate point.
  4. 4. The nano-silver bristle surface characteristic detection system according to claim 1, wherein the specific step of judging whether the macro-uniformity of the gray image is acceptable comprises: Extracting a bristle main body area in a full-surface image of a bristle sample and converting the bristle main body area into a gray image, wherein the pixel gray value of the gray image is positively correlated with the distribution density of surface nano silver particles; Configuring a variation threshold for measuring the dispersion degree of the nano silver particle density distribution and a contrast threshold for evaluating the space distribution regularity, extracting the gray average value and standard deviation of a gray image, calculating a gray variation coefficient, and extracting the image contrast through a gray co-occurrence matrix; If the gray scale variation coefficient is larger than the variation threshold or the contrast is larger than the contrast threshold, judging that the macroscopic uniformity is unqualified and triggering the potential uneven area dividing flow to divide the defect severity, otherwise judging that the macroscopic uniformity is qualified, and entering a depth detection link.
  5. 5. The nano-silver bristle surface property detection system according to claim 1, wherein the specific step of dividing the severity of the defect comprises: carrying out defect positioning on the gray level image, identifying a bare substrate area and a nano silver agglomeration area, recording coordinates of a potential non-uniform area and defect types as agglomeration or deletion, counting nano silver particle density in the potential non-uniform area, and calculating a particle spacing standard deviation; giving the number of the areas with the weight according to the number of the potential uneven areas, and establishing a mapping relation between the number of the potential uneven areas and the weight of the number of the areas; Comparing the density of the nano silver particles in each area with the overall average density, and giving a density deviation coefficient according to the deviation degree; constructing a mapping relation between the defect type and the defect type weight according to the influence degree of the defect type on the bristle performance; and building a multidimensional evaluation model by multiplying the regional quantity weight, the density deviation coefficient and the defect type weight, quantifying the defect degree score, and dividing the defect severity according to the mapping rule of the score and the severity level.
  6. 6. The nano-silver bristle surface property detection system according to claim 1, wherein the specific steps of release detection include: Obtaining a simulation solution containing electrolyte and a fluorescent probe, wherein the fluorescent probe is specifically combined with nano silver particles and generates a fluorescent signal positively correlated with the concentration of nano silver; collecting a simulation solution which is not immersed in a sample as a blank control, and establishing fluorescence substrate data; immersing the bristle sample subjected to friction loss detection into an analog solution, setting an initial acquisition interval by adopting a dynamic time sequence sampling strategy, and adaptively adjusting the acquisition interval according to the real-time fluorescence data change trend; And configuring a change rate threshold and an adjustment step length, calculating the change rate of the fluorescence data in real time, and increasing the acquisition interval according to the adjustment step length when the change rate of the fluorescence data is smaller than the change rate threshold, otherwise, maintaining the current sampling interval.
  7. 7. The nano-silver bristle surface property detection system according to claim 6, wherein the specific step of release detection further comprises: Configuring a sliding window, carrying out sectional analysis on the collected fluorescence data, and calculating the slope of the fluorescence data in the sliding window, judging a release mode according to a preset steep slope threshold and a gradual change slope threshold, wherein if the slope is larger than the steep slope threshold, the release mode is judged to be a release zone leading to the falling of agglomerated particles; And configuring a release stop threshold, counting the number of sliding windows with the slope smaller than or equal to the slow slope threshold when the slope of the fluorescence data in the sliding window is smaller than or equal to the slow slope threshold, and stopping release detection if the number is larger than the release stop threshold.
  8. 8. A nano silver bristle surface characteristic detection method, which is realized based on the nano silver bristle surface characteristic detection system according to any one of claims 1 to 7, characterized by comprising the steps of: Step S1, collecting linear images covering bristle samples, extracting and screening characteristic points of adjacent linear images, and calculating a homography matrix; positioning nano silver particles through template matching and Gaussian fitting, acquiring adjacent linear image offset, and then splicing to generate a full-surface image of a bristle sample; S2, extracting a bristle main body area of the full-surface image, converting the bristle main body area into a gray image, carrying out initial quantitative detection on the distribution uniformity of nano silver on the surface of a bristle sample, and judging whether macroscopic uniformity is qualified or not by analyzing the gray image; Step S3, dividing the severity of the defect if the macroscopic uniformity is judged to be unqualified, and entering a depth detection link including friction loss detection and release detection if the macroscopic uniformity is judged to be qualified; Step 3.1, simulating a friction process of nano silver bristles, performing friction detection on bristle samples with qualified macroscopic uniformity, performing fluorescence imaging on a nano silver particle collecting device after single friction, constructing a nano silver particle distribution image sequence, triggering acquisition of a bristle sample surface damage image by constructing a similarity analysis model of adjacent particle distribution images, constructing a damage image sequence, detecting a damage area by using a reference image difference method, quantifying the area occupation ratio and the spatial distribution of suspected damage areas, judging whether the coating abrasion of the bristle samples enters a stable stage, and stopping friction loss detection; S3.2, immersing the rubbed bristle sample into a simulation solution containing a fluorescent probe when friction loss detection is terminated, collecting fluorescence data of the sample solution, and detecting the release amount of nano silver particles of the rubbed bristle sample in the simulation solution; The specific steps of friction loss detection include: Carrying out fluorescent marking on the bristle sample with qualified macroscopic uniformity, collecting linear images of the bristle sample after fluorescent marking, and splicing to generate a bristle sample reference image; setting a friction working condition, starting a friction test bed to rub a bristle sample, and simultaneously placing a nano silver particle collecting device below a friction area to capture the fallen nano silver particles; After each brushing sample is rubbed for a single time, fluorescent imaging is carried out on the nano silver particle collecting device, nano silver particle distribution images are collected, and a particle distribution image sequence corresponding to the rubbing times is constructed; Performing binarization treatment on the particle distribution image to obtain a nano silver particle mask, and filling cavities in the nano silver particles through morphological operation; Configuring a roundness threshold, analyzing and marking all nano silver particles through a communication area, obtaining the mass center and the area of the nano silver particles, calculating the roundness of each nano silver particle, marking the nano silver particles larger than the roundness threshold as monodisperse particles, otherwise marking the nano silver particles as agglomerated particles; constructing a similarity analysis model of adjacent particle distribution images, extracting the difference of the number of the monodisperse and agglomerated particles and the mask deviation degree of the nano silver particles as characteristic parameters, and obtaining the comprehensive similarity of the adjacent particle distribution images through weighted average; Configuring a similarity threshold, when the comprehensive similarity of adjacent particle distribution images is smaller than the similarity threshold, collecting linear images of bristle samples, splicing to generate surface damage images of the bristle samples, and constructing a surface damage image sequence corresponding to the friction times; detecting damage areas of the surface damage images, configuring damage threshold values based on a reference image difference method, extracting areas with gray value changes larger than the damage threshold values as suspected damage areas, connecting the suspected damage areas through morphological operation, and quantifying the area occupation ratio and the space distribution of the suspected damage areas; Counting the periodic friction times of the last acquired surface damage image, and judging that the coating of the bristle sample is worn into a stable stage and stopping friction loss detection when the periodic friction times are larger than a preset stable detection threshold value; And configuring a limit frequency threshold and a limit area threshold, and stopping friction loss detection when the accumulated friction frequency is larger than the limit frequency threshold or the area occupation ratio of the suspected damaged area is larger than the limit area threshold.

Description

System and method for detecting surface characteristics of nano silver brush hair Technical Field The invention relates to the technical field of visual detection, in particular to a system and a method for detecting the surface characteristics of nano silver bristles. Background Nano silver has been increasingly used in the field of functional bristles, such as oral care brushes, medical cleaning brushes, and the like, due to its unique antibacterial properties. The core performance of the product depends on the distribution uniformity, wear resistance and durability and silver ion release behavior of the nano silver coating on the surface of the brush hair, and the accurate detection of the characteristics is a key for guaranteeing the quality and safety of the product. At present, the detection of the surface characteristics of nano silver bristles faces multiple technical challenges, namely in the aspect of coating uniformity detection, the cylindrical curved surface structure of the bristles causes that full surface coverage is difficult to realize by the traditional planar imaging technology, detection blind areas often appear, and the recognition precision of local defects such as nano silver particle aggregation and substrate exposure is insufficient, so that the quantitative evaluation requirement of coating uniformity cannot be met. In the friction and wear detection link, the existing method mostly adopts static friction tests under fixed working conditions, is difficult to simulate complex and changeable load conditions in actual use, and lacks real-time tracking of the shedding behavior of nano silver particles and the damage evolution of a coating in the friction process, so that the analysis of a wear mechanism has one-sided performance. In the aspect of silver ion release performance detection, the traditional method generally adopts a static soaking or simple oscillation mode to simulate the application environment, ignores the influence of factors such as dynamic shearing force, electrolyte environment and the like in an actual scene on release behavior, cannot accurately distinguish burst release from stable release, and causes the deviation of a release amount detection result and performance under the actual use scene. In summary, the prior art has the problems of incomplete detection coverage, working condition simulation distortion, single mechanism analysis, data association deletion and the like in the detection of the surface characteristics of the nano silver bristles, so the invention provides a system and a method for detecting the surface characteristics of the nano silver bristles in order to overcome the limitations. Disclosure of Invention Aiming at the defects existing in the prior art, the invention aims to provide a nano silver brush hair surface characteristic detection system and method, which aim to solve the technical problems of accurately realizing initial quantitative detection of nano silver distribution uniformity on the nano silver brush hair surface, simulating a friction process in actual use to detect a coating abrasion state, and carrying out differential detection on the release amount of nano silver particles in an abrasion stabilization stage, and realize comprehensive and accurate assessment on brush hair surface characteristics and functional performances through image acquisition processing, multidimensional detection and associated analysis. In order to achieve the above purpose, the present invention provides the following technical solutions: A nano silver bristle surface property detection system comprising: Collecting linear images covering bristle samples, extracting and screening characteristic points of adjacent linear images, positioning nano silver particles through template matching and Gaussian fitting after calculating a homography matrix, obtaining offset of the adjacent linear images, splicing to generate a full-surface image of the bristle samples, extracting a bristle main body area of the full-surface image, converting the bristle main body area into a gray image, and judging whether macroscopic uniformity of the gray image is qualified or not through initial quantitative detection of distribution uniformity of nano silver on the surface of the bristle samples; If the macroscopic uniformity is judged to be unqualified, dividing the severity of the defect, otherwise, judging the macroscopic uniformity to be qualified, and entering a depth detection link, wherein the depth detection link comprises friction loss detection and release detection; The friction loss detection is used for simulating a friction process of the nano silver brush hair, carrying out friction detection on a brush hair sample with qualified macroscopic uniformity, carrying out fluorescence imaging on a nano silver particle collecting device after single friction, constructing a nano silver particle distribution image sequence, triggering acquisition of a brush hair