摘要
针对优质稻米品质国标参数检测的实际需要,在设计稻米图像采集装置的基础上,给出一种基于嵌入式的稻米粒型特性参数测试方法。系统以基于ARM内核的S3C2410为硬件开发平台,以linux系统为软件开发平台,采用USB数字摄像头,结合图像处理与模式识别技术,利用大津法对图像进行阈值分割,通过边缘检测和最大距离方法获得稻米粒型国标参数,测试误差不超过1%。该系统为稻米国标参数的现场测试奠定了基础。
In order to test the quality parameters of high-quality rice,this paper presented a method for testing the rice shape parameters based on embedded system. The equipment used for capturing the image of rice was designed. The hardware platform of the system was S3C2410 with ARM core,and its software development platform was Linux system. A USB digital camera was used,and the technology of image processing and pattern recognition was applied. The threshold segmenting image was completed by using OTSU method. The edge detection method and the maximum distance method were employed to obtain the shape parameters of rice. The testing error was less than 1%. This system lays foundation for the field measurements of national standard rice parameters.
出处
《农业工程学报》
EI
CAS
CSCD
北大核心
2010年第12期131-135,共5页
Transactions of the Chinese Society of Agricultural Engineering
基金
河南省科技攻关项目(.0624100003)