The Research On Identification And Classification Of Cord Fabric Defects

In textile industry,the detection of the defects has been done in the traditional way by workers for various reasons in China. The rapid expansion of computer technology also makes the digital image processing technique develop rapidly. Machine vision technology has become a kind of clever technology and infiltrated in the industry field.Based on the study and research of cord fabric defects detecting technology, we apply machine vision technology instead of hunan vision detection.It can give full scope to machine vision with no fatigue, stabilization, objectivity and high speed.Based on the study and research of the development of the technology, this paper offers the identification algorithm of cord defects based on machine vision. The identification algorithm will be discussed under three headings: cord defects image segmentation, cord defects feature parameters extraction, cord defects identification. Firstly, we carry on the analysis and the explanation to the cord automatic detecting system.Secondly, an approach using minimum error method based on an improved particle swarm optimization is proposed to implement cord defects image segmentation.Thirdly, according to different feature parameters,we choose radial length, zonal length, elongation, euler number as the feature parameters.Finally, the theory of perceptron neural network is introduced and applied to cord defects identification. The cord defects identification algorithm which is proposed has the better effect.