基于网格分解的东巴象形文字分类算法研究

东巴文字作为人类早期的一种向象形文字、标音文字过渡的图画文字形式，既具有图画文字以图表意特点，又具有现代文字使用简单线条表达含义的特点。东巴文字本身的复杂性使其相关研究一直较少且连贯性不强。从东巴文字的构字要素入手，通过分析东巴文字的组成要素、结构特征及造字习惯，给出适用于东巴象形文字的预处理及基于网格分解的分类识别算法。该算法思路简单、复杂度低、易于实现，能够快速实现不同类型东巴文字的检索和识别，具有较好的缩放和平移不变性，从而为东巴文字的造字研究提供强有力的技术支持，也为研究其它象形文字的检索和识别技术提供重要参考。

Dongba Hieroglyphic Classification Algorithm Based on Grid Resolution

Dongba hieroglyph is a kind of very primitive picture hieroglyphs； it not only has a characteristic of pictograph to express meaning by using pictures， but also has a feature of modern word to express the meaning with simple strokes. Because of the complexity of the Dongba hieroglyph itself， the related research on has been limited. In the paper， we start with the structural elements of Dongba hieroglyphs. By analyzing the constituent elements， structural features and writing habits， we give a preprocessing and classification algorithm for Dongba hieroglyphs. The algorithm is simple， low in complexity and easy to implement. It can quickly retrieve and recognize different types of Dongba hieroglyph， and has better scale and translation invariance. The algorithm provides strong technical support for the study of glyphic formation in Dongba hieroglyph and also gives an important reference for the research on the retrieval and recognition of other pictographs.