In order to improve the detection of surface defects of the strip′s p recision and real-time performance,a new optimized QPSO_RBF network is used in strip defect classifica tion and recognition.Firstly,the parameters of RBF hidden layer are determined by using weighted fuzzy C-means (WFCM) algorithm,cluster distribution and evacuation distribution of the strip feature data can be well handled by WFCM algorithm.The algorithm can avoid feature data set equal partition trend,Then, all network parameters are coded to individual particles in this algorithm,the parameters can dynamicaly search optimal-adaptive values in global space by quantum particle s warms optimization (QPSO),the performance of network learning is improved,and th e strip defect classification and recognition expert knowledge base is established.The experimental results show that the algorithm can obtain more excellent network structure,efficient convergence,th e average recognition rate for strip defects is 94.63%,the average course rate is 3.0%,t he recognition of test sample time is 4ms,and the recognition of test sample time is 10ms less than t he each image collection cycle on production.So the algorithm of this paper can provide favorable co nditions for the scene of the high speed production line of steel strip surface defect real-time detection.