摘要
机械式数字仪表的计算机自动抄表在工业和控制领域得到了广泛应用,而仪表图像中字轮的定位是数字仪表自动抄表系统的关键技术。目前,影响机械式数字仪表的字轮定位的问题主要有:表盘图像的背景复杂、类似字轮的干扰信息较多、所采集图像带有随机角度的倾斜。针对以上问题,文中提出了一种基于傅里叶变换的字轮定位方法。该方法首先采用边缘检测和数学形态学处理方法进行有效区域筛选,再利用基于快速傅里叶变换的倾斜校正方法进行表盘倾斜校正,然后采用水平和垂直积分投影法提取字轮区域,最后根据字轮在表盘中的位置特征判断倾斜校正过程中可能出现的图像倒置现象,并进行校正。该方法定位准确、计算简单,有效缩短了系统运行时间,对机械式数字仪表自动读数系统的实时性发展有重要意义。
Automatic reading system for mechanical digital meter has been widely used in industrial and control fields, and the method of print wheel location is the key technology to automatic reading for digital meter. Currently, main problems that affect print wheel location of mechanical digital meter are the complicated background of dial images, interfering information similar to print wheel and the tilting situation with arbitrary angles. Aiming at the above problems, a new algorithm of print wheel location based on Fourier transform was presented. The edge detection and mathematical morphology processing were used to filter effective region. It took advantage of fast Fourier transform to detect tilting angle of dial image and correct the tilting dial image. It accurate- ly extracted the region of print wheel according to horizontal and vertical integral projection operator. The algorithm can locate the print wheel accurately with little computation. It shortens automatic reading system＇s running time effectively and has great significance to the real-time automatic reading system.
出处
《仪表技术与传感器》
CSCD
北大核心
2010年第2期92-94,97,共4页
Instrument Technique and Sensor
关键词
傅里叶变换
倾斜校正
字轮定位
形态学处理
fourier transform
tilt correction
print wheel location
morphology processing
作者简介
张志佳（1974-），副教授，博士，从事图像处理、机器视觉、模式识别和虚拟仪器技术的研究。E—mail：zzjsut@126．com
