Based On Thresholdsegmentation Techniqueof Fabric Defects Detection Research And Implementation

Along with the progress of science and technology, the rapid development of thecomputer technology is widely used in various fields, and the field has undergonerevolutionary changes. In the textile industry, due to the application of newtechnology, weaving machine continuously improve the efficiency of production,production continued to increase, in order to adapt to the growing demand for thematerial. However, in the defect detection process, the traditional manual detection inmany places still exist, due to the physiological factors, lead to artificially slowdetection speed, quality is not stable, has seriously hampered the textile production,become the bottlenecks in the production process. If the application of computertechnology and the automatic detection of fabric, can reduce the human subjectiveconsciousness on the interference, in ensuring the quality of fabric at the same time,greatly improve production efficiency, and the development of China’s textile industryhas important significance.This article through to the domestic and foreign fabric development presentsituation understanding and analysis, in recent years will be common defect detectionmethods such as: mathematical morphology method, gray statistics method, Fu Liyemethod, the wavelet transform method, histogram method and threshold segmentationmethod. The advantages and disadvantages of contrast and analysis, found thethreshold segmentation method has the advantages of simple operation, small amountof calculation, fast calculation speed, improved low cost, wide application range, andcan detect many kinds of characteristic of fabric, so choose it as the study object. Theso-called threshold segmentation technique is actually in a certain threshold valuedivision will target image background image from the separation process. Imagesegmentation method using gray histogram as the research object, based on thehistogram of the transformation, analysis, and thus to obtain a correct threshold,through the threshold on the defect image segmentation, thereby obtaining imagedefects and defect perimeter and area of characteristic informationThe iterative method for a Shuangfeng histogram to get the correct threshold, incomputer vision, the first use of digitized image collection device of digital image,then the image preprocessing, including image, pixel enhancement, denoisingsmoothing operation, make its histogram showing a gray Shuangfeng histogram, and then use the iterative method to obtain a more accurate threshold, can significantlyimprove the effect of image segmentation. For the single peak histogram, throughimproved iteration formula, the gray histogram of iterative transformation, therebyobtaining a basic satisfaction Shuangfeng histogram, and then using the iterativemethod of Shuangfeng map to calculate the threshold, the threshold for imagesegmentation and verified.Through the analysis on the experimental results, it is pointed out that the subjectof some achievements and shortcomings.