摘要
提出一种适于实时在线检测方便面面块破损的方法,即通过建立计算机视觉系统,获取面块图像,针对方便面特点利用"围剿算法"分割出面块图像,然后通过"削切算法"除去面块边缘毛刺,以便于求得面块的"真边界",然后求取面块与其外接矩形面积的比值对方便面的破损进行快速判断,本实验另辟蹊径,其优势十分明显,识别率高,速度快。
An on-line method was proposed to identify damaged instant noodle. Use computer vision to capture the image. Then use ＂around shock method＂ to segment the image. The burr of edge is get rid of after ＂cut processing＂, So ＂real borderline＂ was gainedl Then the area ratio of dough and exterior rectangle was available and made rapid identify on speedy inspection of defaulted. This study method is novelty, superiority, nicety and speedily.
出处
《微计算机信息》
北大核心
2007年第28期238-240,共3页
Control & Automation
基金
江苏省自然科学基金(BK2002005)
国家教委博士点基金(20040299009)
关键词
方便面
形状识别
计算机视觉
图像分割
Instant noodle
Shape recognition
Machine vision
Image Segmentation
