The Research And Design Of Photoelectric Detecting System For Spinneret Plate

Spinneret is a very important component in spinning industry, whether the spinneret shape accord with the standard have direct effect on silk production's quality. At present, Automatic system has been invented in spinneret detecting area,But most of the devices were imported and the price is high.Most domestic fiber factories are still counting on human eye,but human eye-checking process has a systematic flaw.First, in this way, the speed of process couldn't be fast. The workers are easy to get tired too. Second, because of the subjective judgement, the result can't be accurate enough.To solve this problem,overcome the shortage of human eye-checking process improve the efficiency and the precision and improve the efficiency and quality of production in fiber industry we designed this spinneret auto-check system. To improve the automatic procedure of fiber factory and increase the economic benefitsThis graduation thesis discusses the principle and structure of the spinneret inspection system, according to the actual imaging results from the spinneret combined with digital image processing technology, we discussed a practical application for image processing and detection algorithms for estimating. What we did consist of following several points:(1) Picking the microscope and objective lens, according to spinneret to be inspected. To make sure that all of the spinneret can be perfect imaged.(2) Reconstructing the loading platform of the microscope reasonably. To make sure it could provide a better testing environment, shield off the impact of external disturbance. Designing slots to facilitate the search for targets to be detected.(3) Choosing an appropriate CCD to ensure best resolution and speed of response, The CCD should perfectly match with microscope and computer, and provide the image's good quality.With microscope and computer matching to be intact.(4) Coding software to process the spinneret image.In this part we conducted a large number of comparison and used a variety of classical algorithms.We also designed a key algorithm, to ensure the system could automatically achieve the detection..(5) Demarcating the system.To make sure users could calculate the actual diameter of spinneret holes through the microscopic photos.The role of calibration is not limited to detection of diameter of spinneret hole,microscope systems can also be used to measure the size of a variety of small objects.The spinnerets we designed and constructed is based on the machine vision.Now machine vision is used in a variety of industries, like electronics industry, textile industry, wood industry, packaging industry,aviation industry and food processing industry, the types of test targets are endless.We use halogen light under light source,through the microscope, spinneret holes are imaged on CCD target surface,change the optical signal into digital signal, and then passed them to the computer,Then we write the procedures for processing,Obtain the tested results.Through our research and development,the spinneret automatic detection system been successfully designed and constructioned we achieved:these main functions :dynamic testing for spinneret,when we input the type of spinneret,Then ,put The spinneret on the loading platform,click the Start button to automatically detect test to get the number of holes with different blocking degrees.The results will be displayed on the interface,and workers can observed the spinneret image.at the interface simultaneously and the system can keep record of the spinneret that have been tested,We can call the previous inspection records at anytime.Finally, we calibrated the detection system, obtained the actual size that each pixel represents under different magnification So we can get the the actual size and area from the number of pixels calculated.The system is divided into hardware and software two separate parts,about the hardware we selected the Shanghai changfang ZOOM-620E series stereo microscope,It had equipped with 1.3 mega pixel imported CCD,and Lenovo computer to construct the image grabbing and processing system.The software part was developed under the Visual C + + programming environment, combined with the already existing fully-fledged algorithms and the key algorithm personally developed, the software could process the image that hardware captured rapidly and efficiently.According to the software flow pattern, analyzed every step of the following processing procedure compared and chose the best algorithm.Spinneret image preprocessing and morphological identification is the focal point of this article.In the preprocessing part,we used the specific smoothing method to smooth the spinneret image then compared a variety of gray-scale transformation methods,Finally, we selected logarithmic transformation to enhance the gradation of image to detect the edge of the image,we use the sobel operator, and achieved a fine result.Then we used the method to threshold the image obtained binary image.The edge of the image was fairly thick,So We use rosenfeld8 line refinement method to refine the edge.We extracted target area of the original image, and determined whether the area was a circle or not.If the area is circular, then calculate the center of the circle and radius,Multiply with the calibration value to get the actual diameter of the hole.And determine whether the diameter has changed.Then we could know the attrition rate of the determinand.If the target area is not circle, judge it a plugged hole,compare the size with the standard one,detect the plugged percentage ,And put the results on the screen.After the detection, the data should sort by date and saved for future reference. After commissioning and testing the actual speed and accuracy of the system has achieved the desired design requirements. First, once put into commercial use, it can greatly improve the product quality of the chemical fiber industry thus increase the revenue. For Jilin Chemical Fiber as an example, if the quantity and quality of monofilament could completely controlled, and the dyeability could met the European market requirements, at least 1,000 tons of viscose filament yarn could sell to European market in a single year. Bring the company 7.2 million yuan of additional benefit. Second, a single detection system could make profit nearly 45,000 yuan, and the market demand for the system is well below the satisfaction. It could be a 100 million yuan industry. Third, the use of this system can greatly reduce the labor intensity and improve efficiency. It meets the requirement of scientific development. Fourth, the spinning plate manufacturers can use this system to control the quality of the spinneret, improve detecting efficiency and accuracy. Therefore, this project has profound practical significance, can bring substantial economic benefits.