Research On Machine Vision Detection Technology Of Visible Foreign Matter In Bottled Liquid

In twenty-first Century,the rapid development of industrial technology and the continuous improvement of people's living standards have put forward higher demands on the quantity and quality of the products.Detection of visible foreign matters in the conventional method is by artificial detection,like this backward detection methods,there are many defects,such as the detection of slow speed,low precision,easy interference,poor reliability.And it has been unable to meet the high speed and high precision modern production line.Under this social pressure,the appearance of machine vision technology has slowed down this pressure,because its advantages such as non-contact,fast detection speed and high detection accuracy are attracting the attention of manufacturers.In this article,we studied the machine vision detection technology for visible foreign bodies in transparent bottled liquid.A machine vision inspection system for visible foreign objects in bottled liquid is designed.On the basis of reading a large number of references and combined with related technology research,we improve the algorithm to detect moving objects and verify the improved algorithm.the specific research works are as follows:First,the research background and significance of the visual inspection of bottled liquid are introduced.Combining with the production process of bottled liquid products and the application of machine vision technology in the field of industrial detection,the research status of bottled liquid visual inspection technology at home and abroad is analyzed.According to the characteristics of the visible foreign objects in the liquid,the overall design of the visual inspection system for the visible foreign objects in the bottled liquid is proposed.The key technologies,such as light source,lighting mode,industrial camera and so on,are introduced in detail.Secondly,the technology of image preprocessing is studied.First,several common noises in images are introduced.Then,several common filtering algorithms are introduced,and the advantages and disadvantages of each filtering algorithm are analyzed.Based on that,an improved adaptive median filtering algorithm is proposed.The algorithm can not only automatically adjust the size of the filter window,but also carry out the filtering processing according to the difference between the noise point and the signal point.The filtered image is processed by histogram equalization,and the image is enhanced.Thirdly,the detection algorithm of visual inspection of visible foreign objects is studied.First,three common detection methods of moving target are introduced,which are optical flow method,background subtraction method and adjacent frame difference method.On the basis of the adjacent frame difference method,an improved frame difference method is proposed to detect the visible foreign object in this subject,and it can detect the moving target very well.On the basis of symmetric three frame difference method,combined with morphological dilation algorithm,it fills the void well,reduces the probability of void occurrence,and facilitates tracking and recognition of subsequent visible foreign objects.Finally,the tracking and discrimination algorithm of visible foreign objects is studied.First,the kinetic analysis of the black foreign body particles with a density larger than the liquid is carried out,and the motion displacement is exponential in the vertical direction.Then the moving target tracking algorithm based on Mean Shift is used to track the visible foreign objects.A foreign object discrimination algorithm is proposed,which combines the area parameters of the foreign body and the trajectory.Then the software interface of the visual inspection system is introduced.