摘要
针对工业生产中数字式仪表的自动识别问题,利用图像处理技术和匹配技术,提出了一种仪表显示字符的识别方法。通过图像灰度化、直方图增强和中值滤波去噪等技术对图像进行预处理,运用相关匹配和图形模板匹配的方法对输入的字符模式进行初始分类和识别。测试结果表明,算法能够自动、快速、准确地识别出仪表的显示字符。
Proposes an automatic recognition method of characters for the digital instrument display in industrial productions, using the image processing and matching techniques. The image is firstly preprocessed through gray transform, histogram enhancement and mid-value filtering, the input character patterns are initially classified by correlation matching and recognized by graphic template matching. The test results shows that this method can recognize characters of digital instrument automatically, rapidly and exactly.
出处
《现代计算机》
2008年第3期70-72,86,共4页
Modern Computer
关键词
数字式仪表
图像处理
字符识别
模板匹配
Digital Instrument
Image Processing
Character Recognition
Template Matching