Liquid foreign matter visual detection method and system

Abstract

A liquid foreign matter visual detection method comprises the following steps of performing image registration based on Fourier-Mellin transform and an unsupervised deformation registration model; preprocessing the registered image; performing image sequence frame difference on the preprocessed image; post-processing the image after the sequence frame difference; obtaining an adaptive deep learning model; and extracting system flaws by a generalized multi-image matting algorithm. The invention further provides a liquid foreign matter visual detection system, and the precision of image registration is improved. A tiny foreign matter segmentation result is clearer, and background noise is further filtered; the detection precision is ensured, the production benefit is increased, and the production loss is reduced.

Description

technical field [0001] The invention relates to the technical field of machine vision image processing, in particular to a liquid foreign object visual detection method and system. Background technique [0002] Defect detection usually refers to the detection of product surface defects. Surface defect detection uses advanced machine vision detection technology to detect defects such as spots, pits, scratches, color differences, and defects on the product surface. [0003] At present, the use of image processing technology to realize the automation of product quality inspection in the pharmaceutical industry, that is, the intelligent light inspection machine, has been used for many years, but the shape and position of the product inspection objects are all carried out under the condition of constant shape and position. For tasks such as foreign object defect detection in infusion bags, due to changes in object shape, position and external environment, as well as various print...

AI Technical Summary

Problems solved by technology

[0005] 1. The registration accuracy of sequence images is low;

[0006] 2. Apply classic visual inspection to liquid defect detection. Various noises such as reflections and air bubbles have low contrast with tiny foreign objects, making the detection results inaccurate;

[0007] 3. For personnel changes in the operation or fine-tuning of the production formula, some defect fluctuations may be introduced, and even new defects may be added;

[0008] 4. The existing light inspection machine has the function of real-time detection of samples, but lacks the monitoring and alarm of system defects;

[0009] 5. Existing light inspection machines mostly use industrial control computers for detection calculation and storage, which is costly and requires engineers to debug on the spot

Images