基于旋转校正和滑动窗口定位的智能仪表字符识别

摘要

Intelligent instrument has been widely used in various fields of life and production. When it was used with monitoring technology, the problem of measuring dangerous area can be solved efficiently. In this paper, character tilt and location were brought into focus. To solve tilt problem, rotation correction method was proposed based on the theory of perspective projection transformation. To solve location problem, sliding window method was proposed by using gray comparison. Experimental results show that these two methods can solve character tilt and location problem in character recognition of intelligent instrument.%　　智能仪表已经在生产生活的各个领域得到了广泛应用。尤其是通过和远程监控技术的配合，可以有效地解决危险区域工作人员无法进入的难题。在智能仪表字符识别的各项技术中，本文重点研究了字符区域倾斜和字符定位问题。对于倾斜问题，采用了基于透视投影变换理论的旋转校正方法；对于定位问题，设计了基于滑动窗口灰度比对的定位方法。实验结果表明，本文提出的方法有效地解决了倾斜问题和字符定位问题，适用于智能仪表的字符识别。