摘要
随着虚拟现实的不断发展,力触觉装置对纹理图像分类的要求更高了,视觉纹理的不同也会影响力触觉装置的反馈情况。为达到触即有感的临场感,对纹理图像的识别效率提出了更高的要求。基于图像分类理论,提出一种基于局部二值模式(LBP)和灰度共生矩阵(GLCM)的纹理分类方法。对LBP算法进行优化,得到B-CLBP算子,在不增加特征维度的同时提高每一维的数据量,从而得到更精准的分类结果。在此基础上使用SVM分类器对纹理材质进行分类,实验结果表明,该算法在纹理材质分类方面具有良好的性能,并且较传统LBP更加优秀。
With the continuous development of virtual reality,the force haptic device has higher requirements for the classification of texture images.The difference of visual texture should also affect the feedback of the haptic device.In order to achieve a sense of presence,there is a higher requirement for the recognition efficiency of texture images.Based on image classification theory,this paper proposes a texture classification method based on local binary pattern(LBP)and gray level co-occurrence matrix(GLCM).It optimized the LBP algorithm to obtain B-CLBP operator.The data volume of each dimension was increased without increasing the feature dimension,so as to obtain more accurate classification results.Based on this,the SVM classifier was used to classify the texture materials.The experimental results show that the algorithm has good performance in texture material classification,and it is better than the traditional LBP.
作者
陈旭
解天祺
Chen Xu;Xie Tianqi(Nanjing University of Information Science and Technology,Nanjing 210044,Jiangsu,China;Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology(CICAEET),Nanjing University of Information Science&Technology,Nanjing 210044,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2020年第6期242-246,共5页
Computer Applications and Software
基金
江苏省自然科学基金项目(BK20170955)。