摘要
针对目前钢铁物流行业人工计数法存在劳动繁重、效率低下等问题,介绍了一种基于S3C2440硬件平台和嵌入式WinCE操作系统的手持式捆扎棒材计数系统。该系统采用基于连通性和圆形度检测的多模板匹配法来统计棒材根数,可以准确地分辨出每根棒材并实现计数;系统使用CMOS数字图像传感器获得原始的捆扎棒材图像,再由图像处理程序对图像数据进行处理。试验结果表明,系统有效提高了计数准确度,而区域标识和人工修正相配合的方法,使得系统能够在工业现场被广泛应用。
Aiming at the problems that heavy labor and low accuracy of manual counting in logistics of iron and steel industry, a handheld counting system based on $3C2440 hardware platform and embedded WinCE operating system for bundled steel bars is introduced. By adopting the multiple templates matching method based on connectivity and degree of circularity detection, each rod can be precisely identified and counted by the system. Original image of the bundled steel rod is acquired by using CMOS digital image sensor, the image data are processed with image processing program. The experimental result indicates that this method effectively enhances the counting accuracy, and the method combining regional identification and manual correction makes the system widely used in industrial fields.
出处
《自动化仪表》
CAS
北大核心
2013年第4期7-9,13,共4页
Process Automation Instrumentation
基金
国家自然科学基金资助项目(编号:60974097
60904016)
关键词
嵌入式
计数
图像处理
模板匹配
连通性
圆形度
Embedded Counting Image processing Template matching Connectivity Degree of circularity