Abstract
This course design aims to develop and research a handwriting matrix recognition and step-by-step visual calculation process display system, addressing the issue of abstract formulas and complex calculation steps that students find difficult to understand when learning mathematics. By integrating artificial intelligence with visualization animation technology, the system enhances precise recognition of handwritten matrix content through the introduction of Mamba backbone networks, completes digital extraction and matrix reconstruction using the YOLO model, and simultaneously combines CoordAttention coordinate attention mechanisms to improve the accurate grasp of character spatial positions. The calculation process is demonstrated frame by frame through the Manim animation engine, vividly showcasing each mathematical calculation step, helping students intuitively understand the intrinsic logic of mathematical operations. Through dynamically generating animation processes for different computational tasks, the system exhibits high modularity and flexibility, capable of generating various mathematical operation examples in real-time according to student needs. By innovating human-computer interaction methods, it brings mathematical calculation processes to life, helping students bridge the gap between knowledge and understanding on a deeper level, ultimately achieving a learning experience where "every step is understood." The system's scalability and interactivity make it an intuitive, user-friendly, and efficient auxiliary tool in education.
Abstract (translated)
本课程设计旨在开发并研究一种手写矩阵识别与分步可视化计算过程展示系统，解决学生在学习数学时遇到的抽象公式和复杂计算步骤难以理解的问题。通过将人工智能与可视化动画技术相结合，该系统利用Mamba骨干网络提高了对手写矩阵内容的精确识别能力；采用YOLO模型完成数字提取及矩阵重构，并结合CoordAttention坐标注意力机制来提高对字符空间位置的准确把握。使用Manim动画引擎逐帧展示计算过程，生动地呈现每个数学运算步骤，帮助学生直观理解数学操作的内在逻辑。 该系统通过为不同的计算任务动态生成动画流程，在保持高度模块化和灵活性的同时能够根据学生的需要实时生成各种数学操作示例。通过创新人机交互方法，使数学计算过程更加生动有趣，有助于学生在更深层面建立知识与理解之间的桥梁，最终实现“每一步都能理解”的学习体验。系统的可扩展性和互动性使其成为一种直观、用户友好且高效的教育辅助工具。
URL
https://arxiv.org/abs/2505.03800
