针对蒙古族传统家具纹样研究不够深入，也未与现代数字图像处理技术相结和，同时，计算机数字图像处理技术的不断发展为研究蒙古族传统家具纹样的数字化分类提供了可实现的方法和途径。本研究利用数字图像处理技术系统地研究蒙古族传统家具纹样的数字化分类。通过提取主题特征、低维目标属性特征和快速稀疏编码的中间特征，筛选出最优的蒙古族传统家具纹样的数字化特征，采用基于语义分析与融合的数字化分类、基于稀疏与低秩的数字化分类，基于张量流形学习的数字化分类技术对蒙古族传统家具纹样进行数字化分类研究，建立蒙古族传统家具纹样的自动化分类体系。本研究拓展了数字图像处理技术的应用领域，并推动蒙古族传统家具纹样与数字图像处理技术两学科间相互交叉、深入发展，为蒙古族传统家具纹样的自动化分类和开发再利用提供新的思路和理论依据，同时，能更好地利用计算机对蒙古族传统家具纹样进行保护和传承，实现自动化再利用。

The research expands the application fields of digital image processing technique, and promote the fusion and development of the two subjects, Mongolian traditional furniture pattern and digital image processing technique, introduces a new thought and theoretical foundation of Mongolian traditional furniture automation classification and development. Meanwhile, using computer technology, Mongolian traditional furniture pattern will be better preserved, inherited and automation reused. Using digital image processing technique, this research made digitized classification on Mongolian traditional furniture patterns systematically. Getting efficient topic features, low dimensionality object bank and intermediate characteristic of fast sparse coding, to screen out the optimal digital character of Mongolian traditional furniture patterns. Based on the digital classification techniques of semantic analysis and fusion, sparse representation and low rank, tensor manifold learning, promote the digitized classification and establish the automation classification system of Mongolian traditional furniture patterns.Studies on Mongolian traditional furniture patterns no depth development, and no touch into the digital image processing technique simultaneously, the developing of computational digital image processing technique provide a reasonable method to study the Mongolian traditional furniture patterns.