基于LVQ神经网络的青年女性胸部识别模型构建OA

In order to improve the accuracy of young women's chest morphology classification and fill the de-fects of the bra number classification system,this paper constructs a young women's chest recognition model based on LVQ neural network by combining the young women's chest body shape features.The study uses non-contact laser 3D technology to collect a total of 216 young female college students'chest data,and the nine chest feature indicators extracted by factor analysis are clustered by K-means clustering method,and the K value is determined by elbow diagram and contour coefficient diagram,and the chest type is finally classified into four categories.On this basis,the LVQ neural network chest type recognition model was constructed,and the LVQ neural network was trained with the 9 chest feature indicators as input and 4 chest types as output.The re-sults show that the model is trained and tested with a recognition accuracy of 95%and a Kappa coefficient of 0.932.Compared with the BP and PNN neural network models,the LVQ neural network model significantly out-performs the other two neural networks in terms of computational efficiency,model accuracy and stability.