低对比度下的智能电表液晶屏外观检测研究  

Research on LCD screen appearance detection of smart meter in low-contrast

作　　者：谢钢[1] 林茂松[1] 代彦辉[1] 

机构地区：[1]西南科技大学信息工程学院,四川绵阳621010

出　　处：《计算机工程与应用》2014年第2期247-251,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61072138);国防工业技术发展项目(No.B3120110005);四川省苗子工程资助项目(No.2011022)

摘　　要：为了解决智能电表(Smart Meter)自动化检测系统中的关键技术问题,提出了在低对比度环境下智能电表液晶屏外观质量检测的处理方法。实现液晶屏区域的准确分割;采用基于小波变换的图像增强算法在低对比度环境下突出液晶屏上的字符信息;利用形态学梯度对字符信息进行重构,最终实现对液晶屏的外观质量检测。实验结果表明,该处理方法能够有效地解决低对比度环境下的智能电表液晶屏外观质量检测问题,检测可靠性高,具有较高工程应用价值。In order to solve the key technical problem in the smart meter automatic detection system, the smart meter LCD screen appearance quality detection approach in the low-contrast environment is proposed. It achieves accurate seg- mentation of the LCD screen area, enhances the character information of LCD screen based on wavelet transform in the low-contrast environment, reconstructs the character information with morphological gradient, and finally fulfills the LCD screen appearance quality detection. The experimental results show that such treatment is an effective solution to the smart meter LCD screen appearance quality detection in low-contrast environment with high detection reliability, so it has a high engineering applications value.

关 键 词：图像分割 图像增强 形态学梯度 外观检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]