CN-121616604-B - Multi-parameter visual on-line detection method and equipment suitable for barb suture line

CN121616604BCN 121616604 BCN121616604 BCN 121616604BCN-121616604-B

Abstract

The invention belongs to the technical field of online detection, and relates to a multi-parameter visual online detection method and equipment suitable for barbed sutures. And adjusting suture detection parameters according to the parameter adjustment data to obtain target detection parameters, collecting multi-view image data of the current barb suture, and processing by using a multi-view detection method to obtain multi-dimensional detection data. Judging whether the multi-dimensional detection data meet a preset threshold value according to the target detection parameters, if yes, outputting a qualified judgment result, and otherwise, marking the defect position type. The invention can adapt to the detection requirements of multi-material multi-specification barb suture lines, reduce the equipment input cost and improve the detection precision and stability.

Inventors

BAO CHENGCHENG
PAN GUOLIANG
CHEN XIN
WANG KEXIN

Assignees

杭州爱普医疗器械股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260203

Claims (10)

1. The multi-parameter visual on-line detection method suitable for the barb suture line comprises the following steps: Collecting material characteristic data and specification parameters of the barbed suture, formulating suture detection parameters in combination with target detection precision requirements, and obtaining parameter adjustment data by adopting an imaging calibration method according to material reflection characteristics and morphological characteristic analysis; Adjusting the suture line detection parameters according to the parameter adjustment data to obtain target detection parameters, collecting multi-view image data of the current barb suture line, and processing by using a multi-view detection method to obtain multi-dimensional detection data; Judging whether the multi-dimensional detection data meet a preset threshold value according to the target detection parameters, if so, outputting a qualified judgment result, otherwise, marking the defect position type; and analyzing the association relation between the defect position type and the suture line detection parameter to obtain defect detection association correction data, and adjusting the target detection parameter to obtain an accurate detection parameter.
2. The multi-parameter visual on-line inspection method for barbed sutures according to claim 1, wherein the step of formulating the suture inspection parameters comprises: collecting optical characteristics, physical characteristics and processing adaptation characteristics of the suture line as the material characteristic data, and collecting line body parameters and barb parameters as the specification parameters; establishing an accuracy threshold according to geometric parameters, defects and forms of the detection object, and determining the detection efficiency requirement by combining the production speed; Determining resolution, shooting frequency, visual field and focal length, light source type and illumination intensity according to the detection efficiency requirement to obtain image acquisition parameters; selecting a denoising parameter, an edge enhancement parameter and an illumination correction parameter as processing process parameters according to the precision threshold and the target defect characteristics, dividing stations according to preset detection logic, and setting a camera triggering interval by taking a main encoder pulse of a production line as a reference to obtain detection flow parameters; And according to the historical detection parameters and the effect extraction correction logic and the process characteristic correction parameters of the similar products, adjusting the image acquisition parameters, the processing process parameters and the detection flow parameters to obtain the suture line detection parameters.
3. The method for multi-parameter visual on-line inspection of barbed sutures according to claim 1, wherein the step of analyzing the parameter adjustment data comprises: extracting average reflection rate, reflection distribution uniformity and defect-background gray level difference aiming at the material optical properties of the barb suture line as the material reflection characteristics; The method comprises the steps of taking clear extraction of image characteristics as a target, judging whether the reflective characteristics of the current material meet detection requirements or not, and identifying deviation types to obtain reflective characteristic deviation; Extracting the sharpness of the barb edge, the uniformity of the line body shape, the uniformity of barb distribution and the density of surface texture of the barb suture line as the morphological characteristics, and judging whether to influence parameter measurement and defect identification to obtain morphological characteristic deviation; based on the reflection characteristic deviation and the morphological characteristic deviation, correspondingly matching the adjustment direction and the adjustment parameter of the suture line detection parameter; and calculating the adjustment coefficient of the adjustment parameter according to a preset precision requirement and the adjustment direction to obtain the parameter adjustment data.
4. The method for multi-parameter visual on-line inspection of barbed sutures according to claim 1, wherein the step of adjusting the target inspection parameters comprises: The suture line detection parameters are in one-to-one correspondence with the parameter adjustment data, and an adjustment parameter mapping table is established; dividing priorities according to the influence substantivity and response speed of different adjustment parameters on the detection effect to obtain parameter adjustment priorities; And step-by-step executing adjustment operation according to the parameter adjustment priority and the adjustment parameter mapping table, checking whether the adjustment amplitude of each adjustment parameter meets the standard or not by taking the target detection precision requirement as a constraint, outputting the target detection parameter if the adjustment amplitude of each adjustment parameter meets the standard, and otherwise, correcting the adjustment amplitude.
5. The multi-parameter visual on-line inspection method for barbed sutures according to claim 1, wherein the step of obtaining the multi-dimensional inspection data comprises: Calibrating the multi-view camera array according to the target detection parameters, setting a camera acquisition interval, dividing a detection area according to a suture line structure, performing image acquisition, and storing according to a product number, the detection area, a camera number and a time stamp to obtain multi-view image data; carrying out graying, denoising and illumination correction on the multi-view image data, and carrying out geometric correction and region clipping to obtain preliminary processing image data; and carrying out feature extraction on the preliminary processing image data to obtain outline defect features, and carrying out geometric dimension, morphological precision and feature surface defect parameter calculation to obtain the multi-dimensional detection data.
6. The multi-parameter visual on-line inspection method for barbed sutures according to claim 1, wherein the step of determining whether the multi-dimensional inspection data meets a preset threshold comprises: Setting a judgment standard threshold value according to the parameter type, the detection area and the defect level layer by layer based on the target detection parameter; and comparing the multi-dimensional detection data with the target detection data one by one according to the judgment standard threshold value, and outputting the qualified judgment result.
7. The multi-parameter visual on-line inspection method for barbed sutures according to claim 1, wherein the step of marking the type of defect location comprises: Determining defect coordinates based on the multi-dimensional detection data, marking positions of defects by combining the divided detection areas, associating images corresponding to camera visual angles, and marking defective image areas; Classifying the defect image areas according to geometric parameter classes, surface defect classes and morphology classes, and determining defect image types; Marking the grade of the defect image according to the defect size and the influence degree of the defect image area, and classifying according to the product number-disqualification-time stamp to obtain the defect position type.
8. The multi-parameter visual on-line inspection method for barbed sutures according to claim 1, wherein the step of obtaining the defect detection-associated correction data comprises: carrying out standardized treatment on the defect position type to obtain standard defect data, extracting suture line detection parameters of a corresponding defect product, and corresponding to the standard defect data according to a product number and a detection time stamp; Associating the defect type with the parameter type, associating the defect position with the parameter effective range, and associating the defect index with the parameter value deviation to obtain an associated analysis dimension; extracting parameter value distribution characteristics, defect occurrence rate association characteristics and index parameter correlation from the standard defect data as association characteristics according to the association analysis dimension; Calculating sensitivity by taking the defect improvement rate as a sensitivity index, quantifying the influence degree of each suture line detection parameter on the associated characteristic, and screening the parameter reaching a preset sensitivity threshold as a high-sensitivity parameter; And formulating parameter correction rules based on the high-sensitivity parameters and the correlation characteristics to obtain the defect detection correlation correction data.
9. The multi-parameter visual on-line inspection method for barbed sutures according to claim 8, wherein the step of adjusting the resulting precise inspection parameters comprises: Based on the target detection parameters, the target detection parameters and the defect detection association correction data are in one-to-one correspondence, a bidirectional mapping relation is established, and scene correction data are obtained through screening according to the currently detected product materials, specifications and defect types; Setting a correction priority based on the parameter adjustment priority and the influence degree, and calculating an adjustment parameter value corresponding to the correction priority according to the scene correction data; and a conflict coordination strategy is formulated according to a parameter detection result and a device boundary, and the target detection parameter is adjusted according to the adjustment parameter value to obtain the accurate detection parameter.
10. A multi-parameter visual on-line inspection apparatus for barbed suture, applied to the multi-parameter visual on-line inspection method for barbed suture according to any one of claims 1 to 9, wherein the inspection apparatus comprises: The parameter intelligent optimization module is used for collecting material characteristic data and specification parameters of the barbed suture, formulating suture detection parameters in combination with target detection precision requirements, and obtaining parameter adjustment data by adopting an imaging calibration method according to material reflection characteristics and morphological characteristic analysis; the online image detection module is used for adjusting the suture line detection parameters according to the parameter adjustment data to obtain target detection parameters, collecting multi-view image data of the current barb suture line, and processing the multi-view image data by using a multi-view detection method to obtain multi-dimensional detection data; The defect judging and marking module is used for judging whether the multi-dimensional detection data meet a preset threshold value according to the target detection parameters, if so, outputting a qualified judging result, and otherwise, marking the defect position type; And the parameter closed-loop correction module is used for analyzing the association relation between the defect position type and the suture line detection parameter to obtain defect detection association correction data, and adjusting the target detection parameter to obtain an accurate detection parameter.

Description

Multi-parameter visual on-line detection method and equipment suitable for barb suture line Technical Field The invention belongs to the technical field of online detection, and particularly relates to a multi-parameter visual online detection method and equipment suitable for barbed sutures. Background The barbed suture is used as a key consumable in medical operation, and parameters such as diameter precision, barb shape, surface integrity and the like of the barbed suture directly influence operation safety and healing effect, and medical-grade severe requirements are provided for detection precision. However, the existing multi-parameter visual on-line detection method faces a bottleneck in practical application, and severely limits the detection reliability and efficiency. On the one hand, the diversity of barb suture material leads to detection precision unstable. Materials such as nylon, polylactic acid, silk and the like widely used in clinic have obvious differences in reflectivity, transparency and texture characteristics, wherein the high-reflectivity material is easy to generate facula to cover defects, the transparent absorbable material is easy to generate background penetration to cause fuzzy contour, and the high-texture material is easy to be confused with surface defects. The traditional detection method mostly adopts fixed parameter configuration, and can not dynamically adapt to the optical and physical characteristics of different materials, so that the defect omission rate and the false detection rate are high, the critical dimension measurement error exceeds the allowable range, and the high-precision detection requirement of medical consumables is difficult to meet. On the other hand, the problem of inefficiency in multi-view data fusion is prominent. The barbed suture line has a complex structure, needs to cover the needle point, the needle body, the barbed inner ring/outer ring and other multi-region detection, and commonly adopts a multi-camera independent acquisition mode in the prior art, but lacks a uniform synchronous trigger mechanism and data integration logic, so that image acquisition at different visual angles is asynchronous, and parameter measurement references are inconsistent. Meanwhile, a detection blind area is easy to form in a single view, and information redundancy or conflict can be caused in multi-view data fragmentation processing, so that the detection efficiency is reduced, and the accuracy of defect positioning and parameter calculation can be influenced by data fusion deviation. In addition, most of the existing detection methods are in static parameter setting modes, detection parameters cannot be dynamically optimized according to material change and defect feedback, and instability of detection accuracy is further aggravated. Therefore, a multi-parameter visual on-line detection method and equipment suitable for the barbed suture are needed, the pain point in the prior art is solved, and the accuracy, stability and high efficiency of the barbed suture detection are ensured. Disclosure of Invention In order to make up for the defects of the prior art, the invention provides a multi-parameter visual on-line detection method and equipment suitable for barb sutures. The invention is mainly used for solving the problems of unstable detection precision and low efficiency of multi-view data fusion caused by material difference of the existing barb suture line. The technical scheme adopted by the invention for solving the technical problems is that the multi-parameter visual on-line detection method suitable for the barbed suture provided by the invention comprises the following steps: and collecting material characteristic data and specification parameters of the barbed suture, formulating suture detection parameters in combination with target detection precision requirements, and obtaining parameter adjustment data by adopting an imaging calibration method according to material reflection characteristics and morphological characteristic analysis. And adjusting suture detection parameters according to the parameter adjustment data to obtain target detection parameters, collecting multi-view image data of the current barb suture, and processing by using a multi-view detection method to obtain multi-dimensional detection data. Judging whether the multi-dimensional detection data meet a preset threshold value according to the target detection parameters, if yes, outputting a qualified judgment result, and otherwise, marking the defect position type. And analyzing the association relation between the defect position type and the suture line detection parameter to obtain defect detection association correction data, and adjusting the target detection parameter to obtain the accurate detection parameter. The invention provides a multiparameter vision online detection method suitable for barb suture lines, which comprises the following steps of: the optical property,