Research On Machine Vision Inspection System In Umbrella Pole Processing

In the automatic processing of umbrella rods,it is first necessary to turn the front pole to a fixed position for subsequent processing.This process is called the “correction” process of the umbrella rod.At present,the mechanical "correction" of the umbrella rod is usually realized by using a mold and a structural method on the production line,but the method is prone to problems such as mold loss,jamming and correction errors,and affects the processing efficiency of the umbrella rod.In addition,in the processing of high-end umbrella rods,if the umbrella column surface of the welded seam is used as the subsequent processing surface,there may be problems in the use of the high-end folding umbrella due to the scratching of the surface of the rod member,so the umbrella is affected.Before the rod is processed,the identification of the weld bead surface with or without welds is also required.In order to solve the above problems,this topic uses the machine vision and image processing related technology to realize the visual correction of the umbrella section and the identification of the pillar surface weld by non-contact method.The specific research contents are as follows:(1)This topic takes five types of umbrella rods of different cross-section types commonly used in the market as research objects.According to the problems in the process of processing the umbrella rods,the demand analysis is carried out,and the overall scheme of the umbrella rod detection platform based on machine vision is proposed.(2)The pre-processing study of the umbrella section and the umbrella cylinder image is carried out respectively.According to the traditional filtering method,the hybrid filtering method is proposed.According to the opening and closing morphology operation,a multi-scale morphological method is proposed,which is designed according to the triangular threshold segmentation method.An automatic extraction algorithm for the ROI area of the umbrella cylinder surface.(3)For each type of umbrella section,with the contour of the fitted image as the core,the modified polygon algorithm based on the improved DP algorithm and the angle correction algorithm of the shaped umbrella section based on image fusion and local image analysis are designed.The visual correction of the cross section of the umbrella rod is realized reasonably and effectively.For the identification of the cylindrical weld of the umbrella pole,the pattern recognition method is used to classify the cylindrical sample,and the umbrella column weld model is designed and constructed.(4)According to the overall scheme of the umbrella visual inspection system,the system prototype of the visual inspection of the umbrella pole was designed and built,and the interface program of the visual inspection system based on OpenCV and MFC was designed.Through the experiment of correcting the angle of the umbrella section and the weld of the umbrella cylinder surface,it is found that the test results meet the detection requirements.On the built test sample,the experimental test and statistical test information are obtained.The final experimental results show that the umbrella visual inspection system can realize the visual correction of 30 umbrellas per minute according to the umbrella section,and the average angle error measured is about 0.778°.The detection success rate is above 90%.The recognition rate of the columnar weld of the umbrella rod can reach 98%,the false detection rate is about 2%,the high accuracy rate and the low false detection rate can meet the production line inspection requirements.The system is intelligent and automated in the process of the umbrella pole.There is practical significance in it.