Design And Research Of Darlington Pipe Size Inspection System Based On Industrial Vision

With the advances in modern technology, manufacturing also flourishes. Traditional detection mode has been unable to meet the needs of modern manufacturing.The detection technology of industrial products tends to be in high precision, high speed and networking. After production automation problem solving, the industrybegan to seek an industrial vision-based product testing automation. Industrial Vision detection technology is highly favored by the industry in non-contact, high precision, high speed, stability and other advantages.Against this background, this paper has used Darlington tube which is widely used to do the detection to design the overall detection system. The overall framework of the system is divided into two parts, hardware and software, focusing on the analyzing of three constituent units of the hardware part. According to the detection accuracy of the system and the actual characteristics of Darlington, it has selected the optimum resolution CCD camera, low distortion lens and more targeted lighting units and lighting means. Based on the MATLAB software platform with efficient numerical processing capability and basic requirements of high detection accuracy and stability, system software part has studied and designed overall data processing of the detection system. Then it has researched and discussed the image processing algorithm of the system software part, and has selected the extended hybrid median filter as an image filtering method according to comparing the advantages with disadvantages. An improved method of Ostu is thresholding method. It has chosen sub-pixel edge detection and improved Hough transform methods as edge detection and feature extraction methods according to the differences of size detection and defect recognition of the Darlington tube by improving and comparing several commonly used methods in Image edge detection and feature extraction. The system in the detection accuracy and stability has been effectively improved through a series optimal selection and improved methods.And it has conducted many tests by adopting the designed detection system and KC-CV-X200 Series detection system with 20 Darlington tubes and compare their detection results with the real data todo the degree of polymerization analysis of characteristic distribution curve to verify and design a detection system which can conduct stability testing quickly on Darlington tubes’ size and defect. It concludes that the detection system hasgood prospects because of their high precision.