Research On The Automatic Detection System For The Air Tightness Of Latex Gloves

With the increase in domestic labor costs,some multinational companies have successively moved their production lines to other developing countries.Domestic production activities are also affected to a certain extent by labor shortages and recruitment difficulties.The current production of disposable latex gloves has been roughly automated,but on the inspection side,it still follows the old-style manual observation test.Although this method is highly feasible,but due to personal factors,the efficiency of fixed-point detection is not ideal.After rethinking the original detection steps,this paper conceived a detection method that can be combined with machine vision technology.It can not only replace manual work with computers,but also realize streamlined detection,and its detection efficiency is improved.In this paper,an automatic detection system based on the principle of dynamic target detection and PLC control theory has been specifically studied.First,the function that the system needs to realize and the technology of machine vision is briefly introduced and analyzed,and then a new detection scheme is proposed.Clearly stated the details of the image acquisition module of the machine vision system in the scheme.Secondly,according to the functional requirements,the mechanical structure matching the detection scheme was designed,and four device modules are divided according to its subdivision function.Siemens S7-300 is selected as its control system,and I/O points are allocated according to the detection process.Use the ladder diagram language to write the control program.Finally,bubble detection experiment is used to study the detection effect of gloves.After analyzing the acquired images,we put forward opinions from both micro and macro perspectives.Introduce grayscale algorithm based on global mapping function.After verifying the three methods of sensory comparison,contrast evaluation and edge extraction with the current mainstream gray-scale algorithm,the better Lu’s method is determined as the final grayscale algorithm in this paper.Discuss the noise in the experiment,according to its characteristic hypothesis testing,to find a suitable image enhancement method.In terms of dynamic target recognition,the actual results of the application of the inter-frame difference method and the background difference method in glove detection are compared and analyzed.Based on its characteristics and a summary of the law of the detection result images,a two-step detection process was developed,the coordinates of the regions of interest ROI1 and ROI2 are defined,and the detection algorithms that should be used in different stages and the dynamic area ratio standards that match them are introduced.Application of the two new sequences of video detection process to verify judgment standard.The results show that the method can indeed complete the extraction of dynamic targets,and complete the qualified judgment of the glove air tightness test according to the judgment standard.