【摘要】 针对传统的纹理分类算法在家具板材分类中难以区分同纹理不同颜色样品的问题,以及对有图案干扰的同纹理同颜色的一种样品误分多类的情况,文中采用了基于集成高斯支持向量机的分类识别算法。该方法在彩色图片的HSV颜色空间中提取纹理特征参数,利用高斯混合模型统计出的概率矩阵作为支持向量机的输入参数,且用二进制纠错码将二分类扩展到多分类。试验结果表明,该方法比贝叶斯、高斯模型、支持向量机的方法更有效,比人工神经网络有更高的效率。更多还原

【Abstract】 For traditional texture classification algorithm,it is difficult to distinguish the same texture and different color samples in furniture sheet classification,and misclassification of a sample with the same texture and color with pattern interference. An ensemble Gaussian classification algorithm based on support vector machine(SVM) was proposed.The proposed method extracts the texture parameters in the HSV color space of color image,and uses the probability matrix calculated by the Gaussian mixture model as the input parameter of SVM,and expands the binary classification to multi-classification with binary error correction code. Experimental results show that this method is more effective than Bayesian,Gaussian and support vector machines,and has higher efficiency than artificial neural networks.更多还原