摘要
给出了一种用于一次性白色圆形餐具表面污渍检测算法的设计思路。算法首先利用形态学边界检测算法提取圆形餐具的外轮廓,紧接着以外轮廓所对应的前景像素点作为样本点,利用最小二乘法拟合外轮廓,给出了最小二乘法拟合圆曲线的详细公式推导过程。利用获得的圆心坐标及半径参数,算法进一步将餐具划分为不同的检测区域,再分别对各个区域进行污渍检测。结合算法处理流程,给出了采用VC++6.0进行算法实现时的核心函数说明。最后结合实例说明了检测算法的实际测试效果。
An image processing algorithm for detecting disposable circular tableware surface defects is proposed.Proposed algorithm firstly uses morphological method to extract tableware＇s boundary,and corresponding foreground pixels of this boundary are using by least-squares method to fit a circle curve.Center and radius of the circle can be found and used to segment the tableware image into different detection regions.Then proposed algorithm detects segmented regions separately and finally finds out the defects.Codes are developed using VC＋＋6.0.Major functions are commented in detail.Testing results show that the proposed algorithm acts satisfactory performance.
出处
《电子器件》
CAS
2010年第2期249-252,共4页
Chinese Journal of Electron Devices
基金
校设科研基金项目资助(KYS055609080)
关键词
机器视觉
污渍检测
一次性餐具
最小二乘法
machine vision
defects detection
disposable tableware
least-squares method
